Artificial Intelligence Machine learning Startups Technology

Difficulties of managing a Machine Learning project for a data-scientist

There are many difficulties that a data scientist may face while managing an ML project. Some of these challenges include:

  • Data availability and quality,
  • Feature engineering,
  • Model selection,
  • Model tuning,
  • Deployment and maintenance,
  • Legal and ethical considerations

Let’s see these data-scientists’ challenges in more detail.

Data availability and quality

ML algorithms require large amounts of high-quality data to train on. However, it is often difficult to obtain clean and relevant data, which can hinder the performance of the model.

Data availability refers to the ease with which data can be obtained for a particular ML project. Obtaining high-quality data is often one of the most challenging and time-consuming aspects of an ML project. There are several reasons why data availability and quality can be a challenge:

  1. Limited data: In some cases, there may be very little data available for a particular problem. For example, consider a startup trying to build a recommendation system for a new online marketplace. If the marketplace is just starting out and has few users, it may be difficult to obtain sufficient data to train a reliable recommendation system.
  2. Inaccessible data: Even if the data exists, it may be difficult to obtain. For example, data may be stored in a proprietary format or held by a company that is unwilling to share it.
  3. Data quality: Even if data is available, it may not be of high quality. This can include issues such as missing values, incorrect or inconsistent labels, or data that is not representative of the problem at hand.
  4. Data privacy: In some cases, data may be sensitive and cannot be shared for legal or ethical reasons. For example, personal medical records cannot be shared without proper consent.

Ensuring that sufficient and high-quality data is available is crucial for the success of an ML project, as the performance of the ML model is directly related to the quality of the data it is trained on. If the data is of poor quality or is not representative of the problem at hand, the model is likely to perform poorly.

Feature engineering

Creating features that represent the data in a meaningful way is an important step in the ML process. However, this can be time-consuming and require domain expertise.

Feature engineering is the process of creating features from raw data that can be used to train ML models. It is a crucial step in the ML process, as the quality of the features can have a significant impact on the performance of the model. However, feature engineering can be a challenging task for several reasons:

  1. Domain expertise: Creating features that are relevant and meaningful for a particular problem often requires domain expertise. For example, a data scientist working on a healthcare problem may need to understand the medical context in order to create useful features.
  2. Time-consuming: Creating features can be a time-consuming process, especially if the data is large or complex. It may require significant preprocessing and cleaning, and the data scientist may need to experiment with different approaches to find the most effective features.
  3. Lack of guidance: There is often no clear guidance on how to create the best features for a particular problem, so the data scientist may need to try multiple approaches and use their own judgment to determine what works best.
  4. Curse of dimensionality: As the number of features increases, the amount of data needed to train the model effectively also increases. This can make it more difficult to train a model with many features, as it may require a larger dataset to achieve good performance.

Overall, feature engineering is a crucial but challenging aspect of the ML process, and it requires both domain expertise and creativity to create effective features.

Model selection

There are many different ML algorithms to choose from, and it is often not clear which one will work best for a given problem. This can require extensive experimentation.

Model selection refers to the process of choosing the best ML algorithm for a particular problem. This can be a challenging task for several reasons:

  1. There are many algorithms to choose from: There are many different ML algorithms available, and each one has its own strengths and weaknesses. It can be difficult to determine which algorithm will work best for a particular problem, and it may require significant experimentation to find the best one.
  2. Different algorithms work better for different types of data: Some algorithms are more suitable for certain types of data than others. For example, decision trees are a good choice for data with a categorical response, while linear regression is better for continuous responses.
  3. Algorithms may require different types of input: Some algorithms require that the input data be transformed in a particular way, such as scaling or normalization. This can make it more difficult to compare algorithms, as they may need to be tested on different versions of the input data.
  4. It can be difficult to determine the best hyperparameters: Each ML algorithm has a number of hyperparameters that need to be set in order to obtain good performance. It can be difficult to determine the optimal values for these hyperparameters, and it may require significant experimentation to find the best ones.

Overall, model selection is a crucial step in the ML process, but it can be challenging due to the large number of algorithms available and the need to determine which one will work best for a particular problem.

Model tuning

Even once an algorithm has been selected, there are often many hyperparameters that need to be tuned in order to obtain good performance.

Model tuning refers to the process of adjusting the hyperparameters of an ML model in order to obtain the best performance. Hyperparameters are values that are set prior to training the model and control the model’s behavior. Tuning the hyperparameters of a model can be challenging for several reasons:

  1. There are often many hyperparameters to tune: Some ML models have many hyperparameters that need to be set, and it can be difficult to determine the optimal values for all of them.
  2. It can be time-consuming: Tuning the hyperparameters of a model can be a time-consuming process, especially if the model has many hyperparameters or if the training process is slow.
  3. The optimal hyperparameters may depend on the specific problem: The optimal hyperparameters for a model may depend on the characteristics of the specific problem that the model is being used to solve. This can make it difficult to determine the best hyperparameters in advance.
  4. There may be trade-offs between hyperparameters: Adjusting one hyperparameter may improve the performance of the model in one way, but it may also have negative impacts on other aspects of the model’s performance. Finding the right balance between hyperparameters can be challenging.

Overall, model tuning is an important step in the ML process, but it can be challenging due to the large number of hyperparameters that need to be tuned and the time and resources required to do so.

Deployment and maintenance

ML models often require significant resources to train and serve, and they may need to be retrained as the data distribution changes over time.

Deploying and maintaining an ML model can be challenging for several reasons:

  1. Resource requirements: Training and serving an ML model can require significant computational resources. This can be a challenge if the model is large or if it needs to be served in real-time to many users.
  2. Integration with other systems: In many cases, an ML model will need to be integrated with other systems, such as databases or web applications. This can be a complex process that requires the data scientist to work with developers to ensure that the model is properly integrated and serving predictions as expected.
  3. Retraining: ML models may need to be retrained as the data distribution changes over time. For example, a model that is trained to classify images of animals may need to be retrained if it is later used to classify images of a new type of animal that it has not seen before. Retraining a model can be a time-consuming process, and it may require additional resources and data.
  4. Monitoring: It is important to regularly monitor the performance of an ML model to ensure that it is still working as expected. This can involve monitoring the model’s performance on new data, as well as monitoring the overall system to ensure that it is running smoothly.

Overall, deploying and maintaining an ML model requires careful planning and ongoing effort to ensure that it continues to perform well over time.

Legal and ethical considerations

ML projects can raise legal and ethical concerns, such as bias in the data or the potential for the model to be used in harmful ways. It is important for data scientists to be aware of these issues and address them appropriately.

Legal and ethical considerations can be a challenge in ML projects for several reasons:

  1. Data privacy: ML projects often involve working with sensitive data, such as personal information or medical records. It is important to ensure that this data is handled in accordance with relevant laws and regulations, such as the General Data Protection Regulation (GDPR) in the European Union or the California Consumer Privacy Act (CCPA) in the United States.
  2. Bias in data: ML models can sometimes perpetuate or amplify existing biases present in the data used to train them. For example, a model that is trained on data that is predominantly from a particular demographic group may not perform well on data from other groups. It is important to consider potential biases in the data and take steps to mitigate them.
  3. Fairness: ML models should be fair and unbiased in their predictions. For example, a model that is used to predict loan approval decisions should not discriminate against certain groups of people. Ensuring that ML models are fair can be a challenging task, as it may require carefully designing the model and the training data to avoid biases.
  4. Explainability: In many cases, it is important to be able to explain the decisions made by an ML model. This can be a challenge, as some ML models are difficult to interpret. Ensuring that ML models are explainable is important for accountability and transparency.

Overall, legal and ethical considerations are an important aspect of ML projects, and it is important for data scientists to be aware of these issues and address them appropriately.

E-commerce Marketing

How to Market Cosmetics Products in simple words?

Marketing cosmetics products for an e-commerce website can be a challenging task, but there are several strategies you can use to help promote your products and increase sales. Here are a few suggestions:

  1. Use high-quality product images: One of the most important elements of marketing cosmetics products online is to use high-quality product images that showcase the products in their best light. This can help attract potential customers and give them a better idea of what the product looks like in real life.
  2. Offer detailed product descriptions: Along with high-quality product images, it’s also important to provide detailed product descriptions that clearly explain the features and benefits of your cosmetics products. This can help customers make more informed purchasing decisions.
  3. Offer promotions and discounts: One way to encourage customers to buy your cosmetics products is to offer promotions and discounts, such as free shipping or a percentage off their purchase. This can help make your products more attractive to potential customers.
  4. Use social media to promote your products: Social media can be a powerful tool for promoting your cosmetics products. Use platforms like Instagram and Facebook to showcase your products, share customer reviews, and offer special promotions.
  5. Invest in search engine optimization (SEO): Search engine optimization (SEO) is the process of improving your website’s ranking in search engine results pages. This can help your website appear higher in search results and attract more potential customers. To improve your SEO, you can use keywords in your product descriptions and website content, and optimize your website for mobile devices.

Overall, marketing cosmetics products for an e-commerce website requires a combination of high-quality product images, detailed product descriptions, promotions and discounts, social media marketing, and search engine optimization. By implementing these strategies, you can help promote your cosmetics products and increase sales.

Artificial Intelligence Machine learning Technology

What is Machine Learning (ML) in simple words?

Machine learning (ML) is a type of artificial intelligence that allows software applications to learn from data and improve their performance over time without explicitly being programmed. In other words, it gives computers the ability to learn from experience and improve their performance on a specific task without human intervention.

So, What is Machine Learning (ML) in simple words?

ML has become an important tool for solving complex problems in a variety of fields, including finance, healthcare, and e-commerce. It is being used to develop applications that can analyze large amounts of data, make predictions, and take actions based on those predictions.

One of the key benefits of ML is its ability to process and analyze vast amounts of data quickly and accurately. This is particularly useful in industries such as healthcare, where doctors and researchers need to analyze large amounts of data to identify patterns and make predictions about patients’ health.

Another important benefit of ML is its ability to improve over time. As a computer application processes more data, it can learn from its experiences and improve its performance on a specific task. This means that a machine learning algorithm can become more accurate and more efficient over time, without the need for human intervention.

There are many different types of ML algorithms, and each one is designed to solve a specific problem. Some of the most common types of ML algorithms include:

  • Supervised learning algorithms: These algorithms are used to predict the outcome of a specific event based on input data. For example, a supervised learning algorithm might be used to predict the likelihood of a patient developing a certain disease based on their medical history and other factors.
  • Unsupervised learning algorithms: These algorithms are used to identify patterns in data without being given specific labels or output targets. For example, an unsupervised learning algorithm might be used to identify clusters of similar customers based on their purchasing behavior.
  • Reinforcement learning algorithms: These algorithms are used to train a computer to take actions in a specific environment in order to maximize a reward. For example, a reinforcement learning algorithm might be used to train a robot to navigate through a maze by rewarding it for taking the correct actions and penalizing it for taking incorrect actions.

One of the most well-known examples of ML in action is the development of self-driving cars. Self-driving cars use a combination of sensors, cameras, and other technology to collect data about their surroundings. This data is then processed by a machine learning algorithm, which allows the car to make decisions about how to navigate the roads safely.

Another example of ML is the use of natural language processing (NLP) to develop virtual assistants such as Siri and Alexa. NLP is a type of ML that allows computers to understand and generate human language. This allows virtual assistants to understand and respond to voice commands, making it easier for users to interact with their devices.

Overall, ML is a powerful tool that is being used to solve complex problems in a variety of fields. Its ability to process and analyze large amounts of data quickly and accurately, as well as its ability to improve over time, make it a valuable tool for businesses and researchers alike.

Artificial Intelligence Education Technology

What is Artificial Intelligence (AI) in simple words?

Artificial intelligence, or AI, is a term that is often used to describe machines or software that are capable of intelligent behavior. At its core, AI is a field of computer science that focuses on the development of algorithms and systems that can mimic human cognition, such as learning, problem-solving, and decision-making.

So, What is Artificial Intelligence (AI) in simple words?

One way to think about AI is to imagine a computer program that is able to learn and adapt over time, just like a human being. For example, imagine a program that is designed to play the game of chess. At first, the program may not know how to play chess at all, but as it is fed more and more data about the game, it begins to learn and improve its performance. Over time, the program may become so good at playing chess that it can compete with some of the best human players in the world.

Another way to think about AI is to imagine a machine or robot that is able to perform tasks and make decisions on its own. For example, imagine a robot that is designed to assist with household chores, such as vacuuming the floors or taking out the trash. The robot may be able to sense its environment, move around on its own, and make decisions about which tasks to perform and when to perform them. This type of AI is often called “autonomous” AI, because the machine or robot is able to operate without human intervention.

Overall, AI is a rapidly-growing field that has the potential to revolutionize many aspects of our lives. From improving the accuracy of medical diagnoses to automating routine tasks in factories and warehouses, AI has the potential to improve efficiency, reduce costs, and enhance the quality of life for people around the world.


How to choose your Linux distribution? Ubuntu or Debian or CentOS or …?

Choosing the right Linux distribution can be a daunting task, especially for novice users. Linux is a versatile and powerful operating system, but it comes in many different flavors, known as distributions. Each distribution has its own unique features and capabilities, so choosing the right one for your needs can be challenging.

How to choose your Linux distribution?

One important factor to consider when choosing a Linux distribution is your level of experience. If you are new to Linux, you may want to choose a distribution that is easy to use and has a user-friendly interface. Some popular options for beginners include Ubuntu, Linux Mint, and Elementary OS. These distributions offer a simple and intuitive interface and come with a range of pre-installed software and tools to get you started.

Another factor to consider is the type of tasks you will be performing with your Linux system. If you are a software engineer, you may want to choose a distribution that comes with a range of tools and development environments pre-installed. Some popular options for engineers include Fedora, CentOS, and Debian. These distributions are known for their robust set of software development tools and support for a wide range of programming languages.

It is also worth considering the type of hardware you will be using with your Linux system. Some distributions are optimized for specific types of hardware, such as low-power devices or high-performance servers. If you have specific hardware requirements, you may want to choose a distribution that is optimized for your hardware.

Finally, it is important to consider the level of support and community support available for the distribution you choose. Linux is an open-source operating system, so many distributions have active communities of users and developers who can provide support and advice. It is worth checking online forums and communities to see which distributions are well-supported and have a strong community presence.

In conclusion, choosing the right Linux distribution can be a challenging task, but it is important to take the time to consider your needs and preferences. If you are a software engineer, you may want to choose a distribution that comes with a range of tools and development environments pre-installed, is optimized for your hardware, and has a strong community presence.

My recommendations:

I personally use Debian for most of my needs (AI, GPU servers, Webservers, ML servers, App Deployments, Docker images, …).

For development environments (for myself and my team), we tend to prefer Ubuntu as it provides a nice balance of “user-friendly” UIs and tools with the stability and features of the Debian it runs under the hood.

For office tasks (accounting, assistants, marketing and other teams) I recommend Ubuntu desktop.

Artificial Intelligence

What is ChatGPT?

ChatGPT is a powerful and easy-to-use tool that allows users to have natural and engaging conversations with an AI-powered chatbot. It is based on the latest advancements in natural language processing and machine learning, and it can help users quickly and easily get answers to their questions, find information, and more.

One of the main features of ChatGPT is its ability to understand and respond to natural language. This means that users can type out their questions or statements in plain English (or any other supported language), and ChatGPT will understand and respond accordingly. This makes it incredibly easy to use, even for those who are new to chatbots or AI technology.

Another key feature of ChatGPT is its ability to provide relevant and accurate answers to users’ questions. This is because ChatGPT is trained on a large dataset of information and can access a vast amount of knowledge to provide users with the answers they are looking for. This means that users can trust that the answers they receive from ChatGPT are reliable and trustworthy.

In addition, ChatGPT also offers a range of customization options that allow users to tailor their chatbot to their specific needs. For example, users can train ChatGPT on their own dataset of information, which can help the chatbot provide even more relevant and accurate answers. Users can also customize the chatbot’s personality and appearance, making the conversation more engaging and personal.

Overall, ChatGPT is a powerful and easy-to-use tool that can help users to have natural and engaging conversations with an AI-powered chatbot. Its ability to understand and respond to natural language, provide accurate answers, and offer customization options make it an ideal solution for anyone looking to enhance their chatbot experience.

=> Discover ChatGPT at OpenAI

Blogging E-commerce Technology Web Development

Optimize your CS-Cart store

Speed up CS-Cart
Faster websites get better SEO scores and get better indexes in search engines, a Google employee once said “Website should be fast”, your visitors will agree, trust me.
CS-Cart more salesFaster and optimized eCommerce websites do sell +40% more than slower ones, just because they are faster. You “will” agree once you see the money come ;)  trust me.
In the continuity of my “Website optimization” posts, this time we will see how to optimize CS-Cart, but hints and methods written here are useful for all other platforms, just you will need to adapt, easily.

E-Commerce Optimization

I assume you already know about the ROCK SOLID eCommerce software “CS-Cart”, if not, you are missing BIG!|works sponsored this post and asked me to optimize their online marketplace which is based on CS-Cart.

The outcome? A faster CS-Cart store which loads in less than 5 seconds and gets a performance grade of 94/100 and on some pages 99/100 (check for yourself at Pingdom), this literally “ROCKS”.

UPDATE: Based on my work ;) ,|works made a package that CS-Cart users can use to speed-up their stores. Will be available soon!

Now we will see in this post, how you can optimize your site load times and speed.
A little bit more than “few” technical knowledge is required if you want to read on, but you never know, maybe you will learn on the way ;-)
I will expose different steps and aspects of optimization which are also applicable with other websites as we saw in my previous posts.

Time Matters

From an optimization point of view, even 0.1 second in load time matters so let’s get started.

Optimize Cache, Server side:

CS-Cart integrates a built-in cache, but that needs lots of modifications to be “robust” and “reliable”.
Here are some issues with CS-Cart’s cache:

  • It “kind of” compresses the JavaScript files but doesn’t combine them! So you find your self serving multiple files which is BAD for load time.
  • It does not cache external JS files you include with the {script} tag. OK, this might be “over doing” but it might be interesting as for serving “hosted” Google Analytics JS on CS-Cart…
  • It “combines” the CSS files but does not compress them, so you will be sending the user a quite BIG file; and that is BAD for page speed too.
  • It does not compress nor clean the HTML output, and again, BAD for speed.
  • It does not serve files with the right expires header so no one know when to refresh the content! How a browser is supposed to know what and when to cache?

Well, what to do, is to optimize this caching to the maximum and make the page generation faster and faster.
So for example the same time CS-Cart “combines” CSS files, let’s tell it to “compress” the output CSS file. When it compresses JavaScript files “independently” let’s ask for a combined and compressed output file, in the mean while, cache external JS files to limit multi-domain requests for the user.
Backup AlertThese steps are quite complex, and need core modification, so you must be careful when updating/upgrading your CS-Cart installation to keep a backup of your MODs.


smarty logoIf you don’t know it already, CS-Cart runs Smarty for its template engine. Smarty is basically great from an ease of use point of view, but unfortunately, Smarty v2 is not optimized for speed. Well, Smarty did great on its Smarty v3 on speed optimization, and included multiple cache handlers like eAccelerator, APC and others.


Alternative PHP Cache LogoI have to mention that APC caching improves A LOT your site performance, but, Smarty v2 does not contain an APC cache handler, so I made an APC cache handler for Smarty v2, and thus for CS-Cart 3, which makes your content load directly from “RAM” instead of Hard Drive. This means “very fast”.

CS-Cart Core Optimizers

You can download these “Optimizers” from my Downloads page  odience Market, instructions included.

Optimize Cache, Client side:

To tell the browser which files to cache and which ones to not, your server must send correct “Expires headers”.
Unfortunately CS-Cart does not set theme correctly. This causes your client’s browser NOT to cache and ask for every page element at each request. Seriously, you DO NOT WANT THIS, because, wait for it…. BAD for speed!
You will need to edit your HTACCESS file(s) to send correct “cache” headers to the user directly from your server.

CS-Cart Optimized HTACCESS

You can download a sample of the “CS-Cart Optimized HTACCESS” from my Downloads page and for the full version, check odience Market, instructions included.

Combine JavaScript files:

Merge JS and CSS files
As I wrote before, CS-Cart just “compresses” JavaScript files and it does NOT combine the files. You need to combine the compressed JavaScript files to use less bandwidth and send the file FASTER to the user, because JavaScript files are “VERY DANGEROUS” for page speed.
A JavaScript file  literally blocks all other elements from loading until itself gets fully loaded, so you want it to be “alone” and to finish loading, wait for it …. …. …. FAST!

CS-Cart JS Optimizer

Get the “JavaScript Combine&Compress Smarty output filter”; that I wrote; from odience Market, instructions included.

Compress CSS files:

Compress content
There are lots of great Open Source projects (ex. Minify) which allow great CSS optimization, combination and compression. I don’t know why CS-Cart does not integrate one such library?
Well, all that said, you need to compress CSS files, and more precisely, Minify them to let users receive just what they need to “style”, nothing more nothing less.

Use less files, Use more SPRITES:

CSS Sprites example
CS-Cart uses a lot of small icons for its base skin.
If your current skin is using more than 20 different icons (not product images, just icons, ex. account icon, cart icon, live help icon, menu drop down arrows,…) you should think about combining them into a single file to decrease the number of requests a visitor’s browser makes to your server. This combination reduces significantly your page load time and speeds up your CS-Cart store. So, go ahead  and combine your small icons into one or two bigger image sprites to reduce file requests as these requests are BAD for speed.
There are some free tools which help you automate and simplify sprites generation:

CS-Cart Sprites based Basic skin

Still not finished, but I am working on creating a sprites based “Basic” skin for CS-Cart, help is much appreciated.

Use a CDN for your static contents: Europe Pops
CDN, CDN, CDN! (It stands for Content Delivery Network if your are new :D)
If you got the $$$ to run with the big guys, go with AKAMAI, Amazon and other big Content Delivery Networks.
But! You can set your FREE CDN too! just serve your static content from a different sub-domain (or domain), for exemple:
Serve images from

while serving your mains website from

This allows “parallel” downloads which will increase significantly page speed.
Well, serving JavaScript and CSS files from a CDN is something to think twice about, because CS-Cart generates your CSS and JS files every time you refresh the cache so if you are on a CDN, that would happen slower than you expect.
To make CS-Cart “CDN-compatible” you will need to make few changes to your skin’s files. If you focus on moving only your images to the CDN, it would be quite easy, though you should also think about the Addons and their images.
Basically, by overriding the $images_dir SMARTY variable at the right moment, you can tell CS-Cart to go get the images from a CDN, though some minor modifications should be done on core smarty plugin files to let CSS files compile correctly.
Another great idea to setup a CDN for CS-Cart is to mount your CDN space as a partition (mount point) on your server and symlink your static directories (images, css, js, cache) to the mount point, this reduces all synchronisation lags, and saves you from changing core files.
CDN CS-CartBut, a better idea, is to use a “Mirror CDN” like!

I already tested and the service is great. They offer a 14 days free trial, which lets you know if you are ready for CDN.
I really recommend them because of the number of pops (servers worldwide) and their affordable price ($4.90 /100GB of data, THIS IS VERY LOW!). So, if you need a CDN, sign up for and make your CS-Cart s
ite faster!

Let me explain on how to use a mirror CDN for your CS-Cart ecommerce software:

  1. Open an account on
  2. Create a new CDN
  3. If you need SSL (I recommend it if you have a secure store)
  4. Choose Free SSL (shared)
  5. or choose a custom SSL (ex.: “” for 39$/month)
  6. Let 4 minutes pass by, for the CDN to be setup of course!
  7. Modify your CS-Cart store to serve images (you can do it even for CSS and JS, but not really needed, as you would have less than 5 JS and CSS files; it’s up to you!)
  8. Enjoy your Free CDN! (Yes, as simple as that, weird huh?)


Use a cookie-less domain for static components:

CS-Cart Cookie
Sites setting cookies will send/set the cookie file (0.8 – 1.5 KB) for each requested component (even for static images and css/js files).
These static contents do not require a cookie file, so by using a subdomain, or another domain (like a CDN), you will be able to get this point fixed and make your static components “cookie-less”, i.e. about 1K size reduction per element, so again, read the previous section about CDN to setup one for your images (at least).

Optimize your server with HTACCESS and PHP.INI tweaks:

Php.ini and Htaccess optimization for CS-Cart
To oil the gears of your server, you might need to have a look at your .htaccess and php.ini (php5.ini) files.
Some stuff you could do:

  • Set Expires headers based on file types,
  • Tell the browser to cache CSS and JS files,
  • Unset ETags,
  • Set Gzip/Deflate compression for your HTML/PHP files,
  • Secure access to your server and store,
  • Secure file requests and prevent file request attacks
  • Make sure downloadable files are downloadable! weird, huh?!
  • Add the “missing” trailing slash in your URLs,
  • Rewrite requests to WWW,
  • Force secure connections through HTTPS
  • Install a Virtualized Software Firewall through HTACCESS to protect yourself from hackers
  • Protect  your server and store from unauthorized queries and requests
  • Speed up the server with the PageSpeed Apache module
  • and so on…

Specially with PHP.INI (PHP5.INI) you could:

  • Set a default FROM address directly on the server to get your emails “delivered”
  • Increase execution times and upload file sizes to optimize performance and ease file uploads
  • Activate ZEND extensions to boost your server
  • and so on…

CS-Cart optimized PHP.INI & HTACCESS

Check the optimized PHP5.INI file (created and tested on’s Shared hosting) and a tweaked (basic) HTACCESS for CS-Cart at my Downloads page.


There is a lot to do for a better web, by optimizing your websites and e-shops, you use less bandwidth and energy. Apart from making better sales and getting better SEO scores, you make a greener world. Do not hesitate about optimization.


Windows Delayed Write Failed – Latest Solution

UPDATE: Latest article: Window Delayed Write Failed – Solutions –> 

A while back, I wrote a post about the “Windows Delayed Write Failed” error message, and I presented ways to resolve the problem. That worked for many people.


Recently, after installing a fresh copy of Windows XP (with SP3), I got the same problem, and I tried everything I knew to solve it, but no luck!! As the problem’s to do with the USB caching and data transfered from the cache to the disk, I went to my favorite search engine again!


The last thing I found, which solved the problem is to use a piece of software from SysInternals (actually aquired by Microsoft) called Cacheset !



Easy Steps to do:

– Download from my Downloads Page,

– Extract it somewhere on your computer,

– Run Cacheset.exe from where you extracted the file,

– Click on “Clear” button to clear the current working cache.

– Set the maximum working cache to either 64 MB or 128 MB,

– Click Apply.

– Buy me a coffee if this worked for you!  (Link is in the sidebar)


I tried it and now, after one week of intensive tests, I didn’t get the Delayed Write Failed error! I suppose that the problem is solved.


Anyway, if it happens even after setting the cache, do the same process of cache setting (to 64 MB or 128 MB) again as Windows has tendency to go back to its initial configurations.


Note1: Test both the cases; 64 MB as cache and then 128 MB, if you get your disk working on 128 MB, so keep it, as it runs faster.

Note 2: When applying the “Working set maximum” value in Cacheset, calculate the number by multiplying the required cache by 1024. example:  64*1024=65536 which is the value you should give.



This is tested and it works! Digg it if useful.


UPDATE : Latest article: Window Delayed Write Failed – Solutions –>