Today's progress:
- Added Bittrex support. Just have to do full verification tests
- Added AnxBTC support. Just have to do full verification tests
- Got a good start on Poloniex support. Will pick back up on this tomorrow.
Running some backend tests tonight with Bittrex and AnxBTC running along with the other API Cache daemons. So far so good as far as system load
Topic: RollerBot - Alpha testing begins soon! - page 5. (Read 18583 times)
This looks awesome! Keep up the great work. Love all the updates. Would be very interested in being in the alpha/beta test group.
Cryptostocks is actually really good. I have some stock over there.
Nothing too exciting today.. Spending some time implementing support for a few more exchanges that I've had on the backburner for a while. At the rate new exchanges have been appearing, it's beginning to look like RollerBot will end up supporting 25-30 exchanges at launch time, which is a great thing for diversity as well as arbitrage opportunities (and profitability in general). I may spend the next couple of days just in this area, getting as many of the new exchanges in that I can.
I'm also looking into adding CryptoStocks support for cryptostock trading.
I'm also looking into adding CryptoStocks support for cryptostock trading.
Nothing too exciting today.. Spending some time implementing support for a few more exchanges that I've had on the backburner for a while. At the rate new exchanges have been appearing, it's beginning to look like RollerBot will end up supporting 25-30 exchanges at launch time, which is a great thing for diversity as well as arbitrage opportunities (and profitability in general). I may spend the next couple of days just in this area, getting as many of the new exchanges in that I can.
I'm also looking into adding CryptoStocks support for cryptostock trading.
I'm also looking into adding CryptoStocks support for cryptostock trading.
I appreciate you even developing this. For people like me trading can be a pain trying to keep everything straight to make a bit of profit.
This is something that will help so much and maybe make it a bit easier to increase what little I have.
Keep up the great work.
This is something that will help so much and maybe make it a bit easier to increase what little I have.
Keep up the great work.
Once I get a user interface slapped on, I will start bringing you guys in slowly. I haven't forgotten about everyone wanting to help Alpha test - it's a very important part of the development process that will really help me test out capacity of the system.
Thanks for all of your support guys - this has been a lot of work and it's been nice being able to share my development on here without having to deal with the typical negativity and accusations you see fly around here sometimes.
Thanks for all of your support guys - this has been a lot of work and it's been nice being able to share my development on here without having to deal with the typical negativity and accusations you see fly around here sometimes.
Once I get a user interface slapped on, I will start bringing you guys in slowly. I haven't forgotten about everyone wanting to help Alpha test - it's a very important part of the development process that will really help me test out capacity of the system.
Thanks for all of your support guys - this has been a lot of work and it's been nice being able to share my development on here without having to deal with the typical negativity and accusations you see fly around here sometimes.
Thanks for all of your support guys - this has been a lot of work and it's been nice being able to share my development on here without having to deal with the typical negativity and accusations you see fly around here sometimes.
very excited about this 
can't wait.
can't wait.
Hello,
would love to beta test this. Have linux / shell experiance. Also potential customer
Thanks!
would love to beta test this. Have linux / shell experiance. Also potential customer
Thanks!
Daily update:
Today was an excellent day of coding. The plugin system is now reworked and can support plugins as a frontend service, a backend service, or any future service that may be added. I now have the backend trade data plugin running, and things are working as expected. I also went ahead and threaded the plugin system. Each plugin (frontend, backend, etc.) will run in it's own thread and have it's own thread controller. This will give me tons of flexibility moving forward. There are still some small additions needed yet, but nothing I can really do much with until I complete the strategy system. As of today, the rework of the backend services is nearly finished (probably around 90% complete). The few bits that still need to be done have to do with auto-scaling features that will allow more Amazon EC2 instances to spin up as needed along with demand.
Performance-wise, the backend services are doing very well. I'm currently running the backend API cache system on 10 exchanges, along with the trade data plugin at about a .25 CPU load. Not too shabby!
Moving forward starting tomorrow, I'll get back into the website/user interface and get back into the guts of the strategy system. Once this is done I will mostly only have to concentrate on website and UI work.
That's about it for now. Things are looking good to meet the March 15th Alpha testing start date!
Today was an excellent day of coding. The plugin system is now reworked and can support plugins as a frontend service, a backend service, or any future service that may be added. I now have the backend trade data plugin running, and things are working as expected. I also went ahead and threaded the plugin system. Each plugin (frontend, backend, etc.) will run in it's own thread and have it's own thread controller. This will give me tons of flexibility moving forward. There are still some small additions needed yet, but nothing I can really do much with until I complete the strategy system. As of today, the rework of the backend services is nearly finished (probably around 90% complete). The few bits that still need to be done have to do with auto-scaling features that will allow more Amazon EC2 instances to spin up as needed along with demand.
Performance-wise, the backend services are doing very well. I'm currently running the backend API cache system on 10 exchanges, along with the trade data plugin at about a .25 CPU load. Not too shabby!
Moving forward starting tomorrow, I'll get back into the website/user interface and get back into the guts of the strategy system. Once this is done I will mostly only have to concentrate on website and UI work.
That's about it for now. Things are looking good to meet the March 15th Alpha testing start date!
That last paragraph speaks volumes as to how many of us feel right now. GoX was nothing but a ripoff and they are trying to figure out not how to give the BTC back to it's clients but keep it for themselves.
I cannot wait to try this out.
I cannot wait to try this out.
Updating the thread for the day
Today I got my standby daemons working well. What they do is slow down the api queries for markets that aren't currently being traded. This lightens the load on the API provider, as well as the Rollerbot servers.
I've also put some work into the plugin system today to support backend plugins, which will run on a separate machine than user/tradebot plugins. The first backend plugin to run will be the tradedata plugin, which will do nothing but look through the API cache for new trade data on each exchange and each market, and put them into the backend database. This allows Rollerbot to have instant access to historical data without having to wait for dozens of API calls to complete.
I've also been putting a lot of thought into the best way to hit the market with Rollerbot. I've decided to go ahead and fire up two more Amazone EC2 instances so that I can launch with every exchange supported (contrary to yesterday's post). The whole MtGox situation is a prime example of why I wanted to build Rollerbot in the first place - to diversify as much as possible so that if one exchange gets hacked or is dishonest, then my holdings are spread out enough to absorb such a blow. To that end, I need to be sure to offer as much diversity as possible when Rollerbot launches. The sooner users can protect their investments (myself included), the better.
Today I got my standby daemons working well. What they do is slow down the api queries for markets that aren't currently being traded. This lightens the load on the API provider, as well as the Rollerbot servers.
I've also put some work into the plugin system today to support backend plugins, which will run on a separate machine than user/tradebot plugins. The first backend plugin to run will be the tradedata plugin, which will do nothing but look through the API cache for new trade data on each exchange and each market, and put them into the backend database. This allows Rollerbot to have instant access to historical data without having to wait for dozens of API calls to complete.
I've also been putting a lot of thought into the best way to hit the market with Rollerbot. I've decided to go ahead and fire up two more Amazone EC2 instances so that I can launch with every exchange supported (contrary to yesterday's post). The whole MtGox situation is a prime example of why I wanted to build Rollerbot in the first place - to diversify as much as possible so that if one exchange gets hacked or is dishonest, then my holdings are spread out enough to absorb such a blow. To that end, I need to be sure to offer as much diversity as possible when Rollerbot launches. The sooner users can protect their investments (myself included), the better.
Updating the thread for the day
Today I got my standby daemons working well. What they do is slow down the api queries for markets that aren't currently being traded. This lightens the load on the API provider, as well as the Rollerbot servers.
I've also put some work into the plugin system today to support backend plugins, which will run on a separate machine than user/tradebot plugins. The first backend plugin to run will be the tradedata plugin, which will do nothing but look through the API cache for new trade data on each exchange and each market, and put them into the backend database. This allows Rollerbot to have instant access to historical data without having to wait for dozens of API calls to complete.
I've also been putting a lot of thought into the best way to hit the market with Rollerbot. I've decided to go ahead and fire up two more Amazone EC2 instances so that I can launch with every exchange supported (contrary to yesterday's post). The whole MtGox situation is a prime example of why I wanted to build Rollerbot in the first place - to diversify as much as possible so that if one exchange gets hacked or is dishonest, then my holdings are spread out enough to absorb such a blow. To that end, I need to be sure to offer as much diversity as possible when Rollerbot launches. The sooner users can protect their investments (myself included), the better.
Today I got my standby daemons working well. What they do is slow down the api queries for markets that aren't currently being traded. This lightens the load on the API provider, as well as the Rollerbot servers.
I've also put some work into the plugin system today to support backend plugins, which will run on a separate machine than user/tradebot plugins. The first backend plugin to run will be the tradedata plugin, which will do nothing but look through the API cache for new trade data on each exchange and each market, and put them into the backend database. This allows Rollerbot to have instant access to historical data without having to wait for dozens of API calls to complete.
I've also been putting a lot of thought into the best way to hit the market with Rollerbot. I've decided to go ahead and fire up two more Amazone EC2 instances so that I can launch with every exchange supported (contrary to yesterday's post). The whole MtGox situation is a prime example of why I wanted to build Rollerbot in the first place - to diversify as much as possible so that if one exchange gets hacked or is dishonest, then my holdings are spread out enough to absorb such a blow. To that end, I need to be sure to offer as much diversity as possible when Rollerbot launches. The sooner users can protect their investments (myself included), the better.
Just checking in while I have a brief moment.
I've done countless tests with HHVM now, and spent a lot of time talking to the HHVM developers. Unfortunately, HHVM doesn't really fit my use case of having a lot going on in parallel - mainly due to the heavy memory overhead of HHVM. As it doesn't support threading yet, I would have to spawn many hhvm instances which is just too resource-intensive. So for now, I'm going to stick to the normal zend PHP interpreter.
I've spent the last few days doing a lot of benchmarking, and it has helped me come up with a plan for moving forward.
Initially, I wanted to launch fully supporting every exchange out there. Unfortunately due to resource allocation costs it would just cost way too much up front to provision the resources needed to process so much data. What I may have to end up doing is launching supporting fewer exchanges so that membership costs help to get the provisions needed to enable the rest of them (I'm thinking of starting with 8-10 of the exchanges enabled, then bringing more in as the userbase grows (which is just basically flipping a switch in the database for each exchange). This will balance my risk to the point of being manageable in the case of a slower-than-expected startup, while still offering way more than the competition.
With the benchmarking out of the way, I can continue forward with the programming, having a much clearer path ahead. If anyone has input on which exchanges they would like to see starting up, it will be taken into consideration.
Back to work!
I've done countless tests with HHVM now, and spent a lot of time talking to the HHVM developers. Unfortunately, HHVM doesn't really fit my use case of having a lot going on in parallel - mainly due to the heavy memory overhead of HHVM. As it doesn't support threading yet, I would have to spawn many hhvm instances which is just too resource-intensive. So for now, I'm going to stick to the normal zend PHP interpreter.
I've spent the last few days doing a lot of benchmarking, and it has helped me come up with a plan for moving forward.
Initially, I wanted to launch fully supporting every exchange out there. Unfortunately due to resource allocation costs it would just cost way too much up front to provision the resources needed to process so much data. What I may have to end up doing is launching supporting fewer exchanges so that membership costs help to get the provisions needed to enable the rest of them (I'm thinking of starting with 8-10 of the exchanges enabled, then bringing more in as the userbase grows (which is just basically flipping a switch in the database for each exchange). This will balance my risk to the point of being manageable in the case of a slower-than-expected startup, while still offering way more than the competition.
With the benchmarking out of the way, I can continue forward with the programming, having a much clearer path ahead. If anyone has input on which exchanges they would like to see starting up, it will be taken into consideration.
Back to work!
Please add me to beta.
Thanks !!
Thanks !!
great logo i like it
Very cool Logo.
I am looking forward to the beta test and to use it.
I am no good at trading this would help so much.
Thanks for the updates.
I am looking forward to the beta test and to use it.
I am no good at trading this would help so much.
Thanks for the updates.
Thanks to the excellent work by Bitcrea, Rollerbot now has a finalized logo (I only changes some of the color scheme to fit my tastes)!
Newest work I did for RollerBot!
Client: RollerBot
Project: Advanced trading platform
Type of logo: Simple, professional, memorable and trendy.
Solvation Logo represents a roller, turn-over, never ending profit. I've used colors and shapes that has associations with robot.
Any feedback is highly appreciated!
Currently I'm working on: 2 projects.
Am I available? YES!
Please write me a PM or email [email protected]
My official graphic design thread: https://bitcointalksearch.org/topic/bitcrea-professional-graphic-design-logo-brand-identity-webdesign-etc-444106
Precise and high quality work guaranteed!
woot!
Awesome job!
Newest work I did for RollerBot!
Client: RollerBot
Project: Advanced trading platform
Type of logo: Simple, professional, memorable and trendy.
Solvation Logo represents a roller, turn-over, never ending profit. I've used colors and shapes that has associations with robot.
Any feedback is highly appreciated!
Currently I'm working on: 2 projects.
Am I available? YES!
Please write me a PM or email [email protected]
My official graphic design thread: https://bitcointalksearch.org/topic/bitcrea-professional-graphic-design-logo-brand-identity-webdesign-etc-444106
Precise and high quality work guaranteed!
Not a whole lot to report today. I was pretty short on time, but the couple of hours I did have went towards testing things out with HHVM.
Unfortunately just a few days ago the HHVM redis support got broken with an update (I'm testing with the nightly builds ATM while waiting for the stable release that includes mysqli support). I spent some time with one of the devs on IRC and the fix will be pushed out in the next day or two - after which I expect to have a successful test of Rollerbot running on HHVM.
Unfortunately just a few days ago the HHVM redis support got broken with an update (I'm testing with the nightly builds ATM while waiting for the stable release that includes mysqli support). I spent some time with one of the devs on IRC and the fix will be pushed out in the next day or two - after which I expect to have a successful test of Rollerbot running on HHVM.
im so excited cannot wait to review this
Jump to: