Well at least you didn't send this one in caps! :-)
I apologize for the delay, but I did not want to respond until I could do so with some level of specificity. That required me to co-ordinate with numerous people inside the organization (or do the CEO on the IBM commercial thing and say "next week" and watch everyone run around).
I will not repeat that I am disappointed with our handling of this issue (uh, I guess I just did). I will commit to three things:
- there will be new client code by the middle of September
- it will be upgradable such that it will not require breaking scripts to install a new version and
- it WILL be platform agnostic.
To say anything more would be to try and excuse or explain, neither of which is relevant. Once again, sorry.
I will have someone contact you directly on your specific issue.
Regards
Robert Macleod wrote:
Hi Elliot,
Are you now able to share the plans for the client code with us ?
I have just tried to install the latest client (2.8.3.2) and having struggled to get it working resorted to the manual :-)
It is bad enough that there is are only detailed installation instructions for the Linux/Unix platform but I was deeply disappointed to find that the entire coverage for the Windows platform is the paragraph below:-
"The client library has not been specifically tested on the Windows platform. Some have noted problems compiling the ciphers. If you are unable to compile Crypt::DES or Crypt::Blowfish, try Crypt::Blowfish_PP (an implementation written in Pure Perl). While slower than the standard Crypt::Blowfish module, Crypt::Blowfish_PP requires no compilation."
I though that the client code was meant to be platform agnostic but the
required modules for HTTPS do not appear to be available for the Windows
platform.
I was bad enough before that the client had stagnated but it now appears to be moving backwards at least for resellers on the Windows platform.
These issues impact my business and make it more and more difficult for me to justify continuing to stay with OpenSRS.
Regards Rob
-----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Elliot Noss Sent: 07 April 2004 12:12 To: Robert Macleod; elliot noss; Todd Jagger Cc: [EMAIL PROTECTED] Subject: Re: Managed DNS Service concerns (tangent)
Yes and no. The issue here is one of form not substance. There are two things.
The first is a simple HR issue. We have recently added a new fellow, Lee Garrison, as VP, Products (separately, we are very excited about Lee joining us. He brings a bit of a different skill set than we have had in the past). The client code migration path, and communicating it, falls into his area of responsibility. He just started on Monday and understands quite well that this is on his plate for quick turnaround.
The second is how much or how little to communicate. Do we talk about what we are planning to do next week or next year? My preference is always to give you guys as much visibility as possible, but the longer the time horizon the bigger and more complicated the communication becomes.
A lot of work has been done on this issue internally. It is just not yet visible.
I don't want to commit Lee to a date on his third day in the office, but it will be soon. Robert, if there is a specific issue that you are wrestling with please speak to your rep or let me know offlist and we will see if we can get you the info you need to make a decision, or the resources to help you with something.
I know this is not the answer you were looking for, but I hope it helps.
Regards
On Wed, 7 Apr 2004 07:05:10 +0100 "Robert Macleod" <[EMAIL PROTECTED]> wrote:
Hi Elliot,
Are you now able to share the plans for the client code with us ?
Regards Rob
-----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of elliot noss Sent: 05 February 2004 22:26 To: Todd Jagger Cc: [EMAIL PROTECTED] Subject: Re: Managed DNS Service concerns (tangent)
Ok. Time permits.
First, I want to talk briefly about client code and RWI. We are, I am,
in complete agreement with the comments made. The one thing that I want
to make sure is clear is that these items are not being ignored. We have spent lots of time and lots of money (little of it well) on these two issues. In any organization some things go well and others go, um, less well. These two are right at the bottom of my personal list. There
have been too many statements unfulfilled on these two items and I do not want to add to the list. I will say folks are working on these things and, especially with client code, I will be unhappy if you do not see a clear plan shortly (within a month?).
Next, with regard to the level of new services. I personally do not agree with the comments below in a couple of ways. When OpenSRS was first introduced it was (quite good) beta code. The original designer chosen did nothing but make pretty pictures for a few months and we had
to switch all of our plans. We had signed contracts and we had a market
moving at a million miles an hour. We pushed this baby bird out of its nest quite early.
The point of all that is that the level of new services launched today is miles ahead of when we first released OpenSRS and 100's of yards ahead of where the first certs release was. IMHO, each release of a new
service is a little better than the previous.
And SO WHAT. None of that matters, because if there is one lesson we have learned and learned well is that we will always learn more in the first six months being in the market than we ever will just talking with customers and whoever for whatever length of time. If you would rather someone else sand off the rough edges so be it. I would suggest that offering a service early provides two HUGE advantages for you. First, it allows you to participate meaningfully in the evolution of the service.
Second, it allows you to learn how to market the service. IMHO, there is
much more impact today in the way a service is marketed (and by this I mean all elements, bundling, pricing, packaging, not just where you stick the link or what keyword you buy) than in its features.
As for your input, there is an extremely high level of involvement with
customers at all stages of planning and design. Don't believe me? Speak
to your rep and ask to get involved. Anyone willing to be demanding is welcome. Many, many of you have participated in surveys, in placeware sessions, in one-on-one conversations with product management and sales
and in the NSE program. Again, this pales in comparison, no matter how well done, no matter how much done, to being in the market.
Lastly, wrt the comment about the process being used in the development
of blogware vs the development of dns, there are some important contextual differences. They are not what I want to comment on though.
Rather I want to highlight that we are learning not only about each new
service, but also about the process of developing new services in general.
Each time we do it we try and do it a little better. We hope to learn a
little more from each introduction. The approach we took with blogware was much different. Your feedback there is helpful. We will try and take the best of each process and leave the worst. It will never be perfect, but our goal is that it just always be a little bit better.
Be assured, with every thing we do your feedback is what drives it. When it doesn't seem that way please remind yourself that you all provide feedback in a number of different ways, meaning no one of you hears it all, and listening to everyone will NEVER mean pleasing everyone. By definition.
Thanks for this. It is a great post.
Regards
Todd Jagger wrote:
Hello,
Pardon me for inserting this extrapolated tangent into this thread, however Rob's statements I think deserve some (more) discussion here as they pertain not only to the DNS service but other Tucows products
as well.
Let me preface by saying each of us has a different business model, needs, goals, customers, etc. What may be critical to one may be unimportant, or even undesired, by another, and vice versa. Tucows can't be everything to everybody and they have a significant challenge
in trying to tailor their offerings and services to a vastly diverse
customer base. While any endeavor has room for improvement I think overall Tucows has done an exceptional job and, above all, stayed consistent with high standards of ethics and
professionalism. In addition they overall seem sincerely interested in what we want.
Sometimes they also appear to ignore what we tell them, but at least
they're listening. ;-)
Whether or not Tucows offers X service or Y product is not the topic
here. Tucows is going to offer the products and services they determine; that is their prerogative, just as ours is whether or not to resell that service or product, or to do business with Tucows at
all.
What concerns me is the development these products are given and the
level at which they are offered. It seems in each case we're given something one or two notches shy of a kick-ass product, and that directly impacts our abilities to sell them to our customers.
To use Rob's example, the DNS service without the ability to configure
TTL.
And the apparent stagnation of the client code interface and RWI. We
resellers have been bemoaning the state of the client code and RWI usability for literally years. The latest word is that new client code is important and probably 6+ months down the road.
I remember
that same "official word" perhaps 2 years ago, back when the SF client
was the model on which the client code was to be built. Specific bugs
and suggestions have gone unimplemented. (Does the client code currently require a payment method - e.g. credit card input - for renewals? This was the first reason I went to the SF code. We don't
want to keep numbers on file, customers' credit cards expire or they change cards or addresses; what is Tucows's model for getting payment
on renewals? None apparently from the client code.)
The email product has the potential to be a great outsourced service
for those of us that offering fits our needs (mine does), and while some of the product are excellent, it falls short of being superior on
multiple levels. The new feature additions are an improvement but don't quite take it to the A list. The webmail interface is still clunky, even compared to something like Squirrelmail, and doesn't even
come close to the web interface of Cyrusoft's SilkyMail. And besides,
how many people want to use webmail for their primary mail client?
Not many I know. My clients want the features to be usable from their
email client, and the webmail is something to use when they're not at
their computer.
Features like the shared address books are great but aren't going to
mean anything to my customers unless they can share them from a mail client. There's no mention of the protocol (is it LDAP?
ISMP? ACAP?
or something proprietary from Stalker?) There are many other issues,
some of which I've raised, and Bruce & Peter, you know how to reach
me.
:-)
The point here is that with core reseller products it seems we're
given
a less than complete, and thus less than competitive, solution. It almost seems that the products (at least the Email and Managed DNS) were determined prior to extensive discussion about what resellers needed/wanted in the offering, and now that the products are out there
it might be difficult or impossible to mold them to what our customers
need. This leaves us in a difficult position of either offering something we're not excited about, or not offering it at all.
This is in stark contrast to what's going on with the Blogware development. That product is being tested, hammered on, Bugzilla'ed,
discussed in detail, and most importantly modified to what the resellers want. I'm convinced it's going to be, already is really, a
top-notch product that's ahead of the curve, not behind it like Email,
Managed DNS, client code, RWI.
IMHO, I'd really like to see Tucows re-tool the quality of the
products
in the same spirit they're developing Blogware. Remember guys, this
is
technology --- you don't want to be playing catch-up.
Thanks for listening tj
-- Elliot Noss Tucows Inc. 416-538-5494 enoss.blogware.com
-- Elliot Noss Tucows Inc. 416-538-5494 enoss.blogware.com
