Best Server type to run a epp dac drop catching script?

Hi all,

I’m in the process of choosing a developer to create a script to poll the real time DAC.

And register any caught domains via epp, should this be run on a windows server?

And should the server be set up in any special way?

Many thanks for help received.

P.s the questions are so i can get a better understanding myself.

 

You'll not get many people giving up the details of their code, but you can code on any OS you like or you have the skill set for.

You should really be asking your developer how they want the server setup as everyone will do things differently.

 

Many thanks for your time in replying...

I do understand what you mean, it’s just a bit of mine field and trying to choose the best solution, because as you stated “everyone will do things differently".

It’s a lot of money for me to get the script created and developed, but you got to be in it to win it.

 

What are the approximate costs out of interest?

 

Hi wb,

Quotes between £1000.00 and £2,500.00 plus vat for a epp system set up on server to poll the real time dac.

With a interface that’s web based where a domain name is uploaded which also allows you to set the polling rate and what hours to poll etc.

The firm that quoted £2,500.00 plus vat do good work, i.e work for leading bank creating web apps and scripts etc.

Does that sound about right?

 

I'm not the best person to ask about costs as I haven't made any enquiries, it sounds like it is probably about right though. I imagine it could be cheaper if outsourced to another country, but not sure how good the quality would be.

Would quite like to get a script made too ready for when I finally get round to joining Nominet and getting a TAG though, fancy sharing the cost? :lol:

 

The only issue I would flag is the real-time element, web app developers generally do not need to understand about sub millisecond timing, they may well optimise web code, but I guess not down to that level.

For that figure, do you get the code so you can mod in the future or do you have to go back to them for changes.

At £2.5K a pop, I'll have to consider creating a GUI & releasing my code

 

Hi Monaghan,

Thank you for your thoughts...

I am hoping to get a script that will poll at rates as low as 85 to 95 mls (not for 23 hours of course, keeping within the real time dac quota) polling one domain.

The code would be realised to me and is to be set up on a server I set up myself, I will be able to choose what hours I poll via a website based interface.

I am going to insure whoever I go with to create the script that I can test it before the final payment is made.

So book a few public drop catching services for the most non competitive domains dropping and set how it fares up (over the course of a week or so).

Then test it on what I belive to be middle of the road competitive level domains.

 

Hi WB,

If i do get a script build and it is very competative I wouldnt sell it....

 

I was just about to post this as a new thread but this seems a good place! £2,500 is a good price, but I think with the number of actual hours involved you will find they won't be able to deliver the script you want.

------------------

In my last thread I was asked how much would it cost to develop a drop catch system?

This is an interesting question and not easily answered. Depending on the route you take the costs will vary wildly. Not only do they vary, but also the quality of the software varies.

This type of software is not a one-off build. The software doesn’t follow to a normal linear client development cycle: spec, build, test, set live, bug fix and pay invoice. It is impossible for any developer to get this software right first time, NO matter how good they are. That is unless they are really lucky, but if you read the thread you will hear that luck plays a big role! It requires the first 5 stages to be repeated, over and over again. And they may even find themselves starting again, for example a different development platform or catching strategy.

As you can read even monaghan has moved from a windows platform to linux.

The two routes for development will be hiring a developer or doing it yourself.
Hiring a developer options:

1. Outsource via an online community
2. Hire a software development company
3. Employ a developer full-time
4. Hire a local freelancer

The problem I see option 1, 2 and 4 is they will not be able quote correctly. As I have tried to point out above the software is not a one off build. And your or their biggest problem is going to be the amount of tweaking and maintenance involved. How can you correctly quote for something that is an unknown? You will either get over charged or find yourself with a piece of software that is useless!

Also this route opens up another issue. You are paying someone to build something that they could then go off and use themselves (even if you employ them). You could even find that your software doesn’t quite perform as well as theirs…nuff said! When they realise the possible profits to be gained, are they really going deliver a competitive product. When you are not a programmer it doesn't take much to introduce the odd sleepy thread.

DIY

If you can program this is probably the best option and will only cost your time or marriage. You will be able to tweak to your heart’s content. If you are a web developer you may find the system challenging as you are going to be building service that run on the operating system. You will need to think about system performance down to the millisecond. This isn't something you'd normally worry about when developing web applications.

You will find yourself waking up to see if the system has run through the night trouble free. You will find yourself tracking system exceptions and monitoring network connections. Trust me finding that your system has fallen over or you lost connection Nominet’s system really doesn’t start the day off good!

My system is still in testing and I am under no illusion that it will go live anytime soon. Every day I learn something new, try another performance tweak, moving closer and closer to my end goal.

-----------------

Jim, if you commission the work, please make sure you put a solid technical specification in place that doesn't leave anything to question. I personally think you will end up spending more than £2,500 for something that is really competitive.

 

Hi Gareth,

Thank you very much for your time in posting... Real food for thought?????

I dont have any tec know how, to lay down the specification - none.

I would be putting my trust into someone else, 4 days was the time scale given for them to create the script...

So it would seem this is impossible then???

 

If you don't have the knowledge, then how can you confirm whether the script is operational or not?

4 days to produce the scripting is not unreasonable, however, 4 days to research the techniques, cut the code and test seems a bit optimistic. I would suggest most catch systems have evolved over many iterations and the chances of getting a good optimised script from someone who's not done it before is quite limited.

Putting on a software developer's hat (I've been directly and indirectly involved with software development and testing for a significant proportion of my working life)

Have you thought about the test criteria?
How will you test this?
To get an "available" response from the DAC is easy, >95% of the names in my list hit the DAC and get an available reply, unfortunately far less than this return a response code of 1000 from the DAC.
How will you determine whether the 2302 response code is network related, too slow in sending the EPP request or whether you happened to poll between someone else getting an available and issuing an EPP create?

I would not want to go with an unproven freelance for a task of this degree if I didn't have a good level of tech skill to define a proper test plan.

A basic bit of code to poll the DAC, determine the reply and fire an EPP request is easy to do, the rest is in implementation and tweaking.

Choose wisely, perhaps take a step back and do a little more research before parting with hard earned cash. Get it right and you'll earn back the cash, get it wrong and you have a white elephant.

 

Alex is 100% right here. And without repeating myself you need do some research before parting with any money. To be honest anyone with experience isn't going charge 4 days work. If you came to me it would cost:

1 day for dac and epp create script
2 days database and web interface
20 days rebuilding and testing!!!!

It is not an everyday job

Good luck

 

:d .........

 

Jim, I sent that message last night whilst drinking, so apologies if I sound negative. The point I am trying to make is, make sure you have a good budget and realise you may have to spend a bit more to get something that is competitive.

 

Hi Gareth,

No apologies what so ever, I’m really grateful for your input and thoughts (and monaghans).

My line of thinking is no drop catcher is really going to give you a copy of their best working script?

But hopefully a good developer (not in the drop catching world) would create a good script and there’s no direct competition.

If you book a slot with some of the great private drop catchers for say £40.00, due to the small amount of money... understandably they will only put a certain amount of resource into that catch? I would of thought (a lot less than say the 432,000 quota for the real time dac).

So i was hoping as a starting point the script might (when polling for one domain at a time) now and again catch a lower to meduim level domain that I would have booked with a public catcher.

Point being I have been using private drop catchers for some time and spent a lot, so this would then pay for the script

 

I think the point Gareth is trying to make is that the initial scripting is just the starting point, tweaking then goes on day by day (probably indefinitely). If you have to pay the programmer each time to make a change, you'll be better off remaining using catchers.

My point was that if you don't understand the tech stuff, how can you test and know if the code is optimal? You risk signing off a pile of poo that may meet the spec you issued, but doesn't do the job. The programmer may have used his favourite .Net or PHP library, but there may be a quicker / better way of doing things. Sadly, real time programming or low level network IO are not common disciplines these days.

 

Jim, we are not trying to scare you off, but these are the realities of this game. I will be up front with you have only just started catching (well trying too). But I have been programming for nearly 11 years now and this has so far been an interesting journey.

 

Wouldn't mind having a look at a drop catching script - does anyone offer old ones which perhaps aren't competitive now?

Was browsing through very old threads earlier on and it would appear that there did used to be a couple of places offering free/paid scripts, but can't find any now. I don't mean I want up to date software which cost thousands, but something just to have a look over to learn how it works.

Hopefully that doesn't sound too much like an unreasonable request

 

I doubt anyone will give away "old code" as the fundamentals don't change (unless there's a big change like the introduction of epp)

The basic code is quite simple though, you just dump your desired name into the DAC, look for a domain.co.uk,n reply and fire off an epp create (this is just an XML string containing your domain, name & address details) and hopefully getting this in before anyone else. Set your EPP & DAC IP's in your Nominet account, fire up your favourite compiler / scripting language and off you go.

The "Secret" bit is how you implement the functionality, which libraries you use, which you code yourself, how you configure your timings and other such "techie" things.