Search Results

Search found 15103 results on 605 pages for 'programmers notepad'.

Page 263/605 | < Previous Page | 259 260 261 262 263 264 265 266 267 268 269 270  | Next Page >

• Naming methods that perform HTTP GET/POST calls?

  - by antonpug

  In the application I am currently working on, there are generally 3 types of HTTP calls: pure GETs pure POSTs (updating the model with new data) "GET" POSTs (posting down an object to get some data back, no updates to the model) In the integration service, generally we name methods that post "postSomething()", and methods that get, "getSomething()". So my question is, if we have a "GET" POST, should the method be called: getSomething - seeing as the purpose is to obtain data postSomething - since we are technically using POST performSomeAction - arbitrary name that's more relevant to the action What are everyone's thoughts?

  Read the article

• Right multi object dependance design

  - by kenny

  I need some help with a correct design. I have a class called BufferManager. The push() method of this class reads data from a source and pushes it to a buffer repeatedly until there is no more data left in the source. There are also consumer threads that read data from this buffer, as soon as new data arrives. There is an option to sort the data before it comes to buffer. What I do right now is that BufferManager, instead of pushing data to the buffer, pushes it to another "sorting" buffer and starts a sorting thread. SorterManager class reads the data, sorts it in files and push()es the sorted data into the buffer. There will be a bottleneck (I use merge sort with files) but this is something I can't avoid. This is a bad design, because both BufferManager and SorterManager push data to a buffer (that consumers read from). I think only BufferManager should do it. How can I design it?

  Read the article

• How to integrate technical line/functional manager into Scrum team?

  - by thegreendroid

  We have recently had a new line manager start who is managing our Scrum team. He is immensely experienced in our field but is relatively inexperienced at Agile/Scrum. He has extensive technical expertise in embedded software (the team's domain) that would go to waste if not utilised properly. However, the team is wary of making a line manager part of the Scrum team. The general consensus is that the line manager should not be part of the Scrum team at all. There are a number of issues that may crop up, e.g. the team may start "reporting" to the manager (i.e. a daily status update!), the manager may start to micro-manage team members etc etc. As it currently stands, he has already said that he feels like an outsider within the team. We really want to make use of his technical skills, we'd be foolish if we didn't because we are a relatively inexperienced and young team of twenty somethings. What would be the best approach to integrate a senior "technical" line manager in a Scrum team and make him feel like he is part of the team?

  Read the article

• How to handle editing a large file for a non-technical user

  - by Luke

  I have a client who is given a tab delimited .txt file containing hundreds of thousands of rows. I have a user story as follows: As a user I want to take the text file and add a new value at the end of each line which contains the concatenated value of two of the columns. for example if the file read text_one text_two I need to output the following (preferably to a .txt file) text_one text_two text_onetext_two My first approach was to ask the vendor supplying the file to do the concatenation before providing the file, the easiest way to solve a problem is to eliminate it right? however they are very uncooperative and have point blank refused. I've looked at building a simple javascript application that does this client side so a non-technical user could select the file using a file selector. This approach has a few problems The file could be over a GB in size and so can't be loaded straight into memory, I've tried and the browser crashes There is no means to write a file in javascript so I'd need to output the content to the screen and have the user save it (somehow) I was thinking if I could get around the filesize limitations I could just output the edited content to the page and have the user save the page as a .txt file, however I think there is a better way than using javascript that will still accommodate the users lack of technical know-how. Please consider this question to be stack agnostic, but bear in mind that a nice little shell script or python script would be deemed unsuitable for a non technical user unless there is a way of "packaging" it nicely for a non-technical user. Updates The file is too large to open in excel. The process needs to be run weekly, but it doesn't require scheduling or automation...(yet)

  Read the article

• Code structure for multiple applications with a common core

  - by Azrael Seraphin

  I want to create two applications that will have a lot of common functionality. Basically, one system is a more advanced version of the other system. Let's call them Simple and Advanced. The Advanced system will add to, extend, alter and sometimes replace the functionality of the Simple system. For instance, the Advanced system will add new classes, add properties and methods to existing Simple classes, change the behavior of classes, etc. Initially I was thinking that the Advanced classes simply inherited from the Simple classes but I can see the functionality diverging quite significantly as development progresses, even while maintaining a core base functionality. For instance, the Simple system might have a Project class with a Sponsor property whereas the Advanced system has a list of Project.Sponsors. It seems poor practice to inherit from a class and then hide, alter or throw away significant parts of its features. An alternative is just to run two separate code bases and copy the common code between them but that seems inefficient, archaic and fraught with peril. Surely we have moved beyond the days of "copy-and-paste inheritance". Another way to structure it would be to use partial classes and have three projects: Core which has the common functionality, Simple which extends the Core partial classes for the simple system, and Advanced which also extends the Core partial classes for the advanced system. Plus having three test projects as well for each system. This seems like a cleaner approach. What would be the best way to structure the solution/projects/code to create two versions of a similar system? Let's say I later want to create a third system called Extreme, largely based on the Advanced system. Do I then create an AdvancedCore project which both Advanced and Extreme extend using partial classes? Is there a better way to do this? If it matters, this is likely to be a C#/MVC system but I'd be happy to do this in any language/framework that is suitable.

  Read the article

• How do you manage a complexity jump?

  - by glenatron

  It seems an infrequent but common experience that sometimes you're working on a project and suddenly something turns up unexpectedly, throws a massive spanner in the works and ramps up the complexity a whole lot. For example, I was working on an application that talked to SOAP services on various other machines. I whipped up a prototype that worked fine, then went on to develop a regular front end and generally get everything up and running in a nice, fairly simple and easy to follow fashion. It worked great until we started testing across a wider network and suddenly pages started timing out as the latency of the connections and the time required to perform calculations on remote machines resulted in timed out requests to the soap services. It turned out that we needed to change the architecture to spin requests out onto their own threads and cache the returned data so it could be updated progressively in the background rather than performing calculations on a request by request basis. The details of that scenario are not too important - indeed it's not a great example as it was quite forseeable and people who have written a lot of apps of this type for this type of environment might have anticipated it - except that it illustrates a way that one can start with a simple premise and model and suddenly have an escalation of complexity well into the development of the project. What strategies do you have for dealing with these types of functional changes whose need arises - often as a result of environmental factors rather than specification change - later on in the development process or as a result of testing? How do you balance between avoiding the premature optimisation/ YAGNI/ overengineering risks of designing a solution that mitigates against possible but not necessarily probable issues as opposed to developing a simpler and easier solution that is likely to be as effective but doesn't incorporate preparedness for every possible eventuality?

  Read the article

• Write own messaging system vs. utilize existing ones

  - by A.Rashad

  We are trying to have our own startup, with a middleware application to glue small applications with enterprise legacy systems. for such middle-ware to function properly, we will need some sort of messaging system to make different components talk to each other in a reliable way. the alternatives are: use an existing messaging system, such as 0MQ, jBOSS, WebSphere MQ, etc. build our own messaging system the way we see the problem I am more biased towards the later option for the following reasons: to have more control over our final product to avoid any licensing problems later on to learn about messaging while writing the code to invent something new, that might cost us lots of $$$ if reused an existing system What would you do if in my shoes?

  Read the article

• Is there a theory for "transactional" sequences of failing and no-fail actions?

  - by Ross Bencina

  My question is about writing transaction-like functions that execute sequences of actions, some of which may fail. It is related to the general C++ principle "destructors can't throw," no-fail property, and maybe also with multi-phase transactions or exception safety. However, I'm thinking about it in language-neutral terms. My concern is with correctly designing error handling in C++ functions that must be reliable. I would like to know what the concepts below are called so that I can learn more about them. I'm sorry that I can't ask the question more directly. Since I don't know this area I have provided an example to explain my question. The question is at the end. Here goes: Consider a sequence of steps or actions executed sequentially, where actions belong to one of two classes: those that always succeed, and those that may fail. In the examples below: S stands for an action that always succeeds (called "no-fail" in some settings). F stands for an action that may fail (for example, it might fail to allocate memory or do I/O that could fail). Consider a sequences of actions (executed sequentially from left to right): S->S->S->S Since each action in the sequence above succeeds, the whole sequence succeeds. On the other hand, the following sequence may fail because the last action may fail: S->S->S->F So, claim: a sequence has the no-fail (S) property if and only if all of its actions are no-fail. Now, I'm interested in action sequences that form "atomic transactions", with "failure atomicity," i.e. where either the whole sequence completes successfully, or there is no effect. I.e. if some action fails, the earlier ones must be rolled back. This requires that any successfully executed actions prior to a failing action must always be able to be rolled back. Consider the sequence: S->S->S->F S<-S<-S In the example above, the first row is the forward path of the transaction, and the second row are inverse actions (executed from right to left) that can be used to roll back if the final top row actions fails. It seems to me that for a transaction to support failure atomicity, the following invariant must hold: Claim: To support failure atomicity (either completion or complete roll-back on failure) all actions preceding the latest failable (F) action on the forward path (marked * in the example below) must have no-fail (S) inverses. The following is an example of a sequence that supports failure atomicity: * S->F->F->F S<-S<-S Further, if we want the transaction to be able to attempt cancellation mid-way through, but still guarantee either full completion or full rollback then we need the following property: Claim: To support failure atomicity and cancellation mid-way through execution, in the face of errors in the inverse (cancellation) path, all actions following the earliest failable (F) inverse on the reverse path (marked *) must be no-fail (S). F->F->F->S->S S<-S<-F<-F * I believe that these two conditions guarantee that an abortable/cancelable transaction will never get "stuck". My questions are: What is the study and theory of these properties called? are my claims correct? and what else is there to know? UPDATE 1: Updated terminology: what I previously called "robustness" is called atomicity in the database literature. UPDATE 2: Added explicit reference to failure atomicity, which seems to be a thing.

  Read the article

• Serialized values or separate table, which is more efficient?

  - by Aravind

  I have a Rails model email_condition_string with a word column in it. Now I have another model called request_creation_email_config with the following columns admin_filter_group:references vendor_service:references email_condition_string:references email_condition_string has many request_creation_email_config and request_creation_email_config belongs to email_condition_string. Instead of this model a colleague of mine is suggesting that strong the word inside the same model as comma separated values is efficient than storing as a separate model. Is that alright?

  Read the article

• Will dolphins die if I use REST "as CRUD"?

  - by l0l0l0l0l

  Recently I moved to Laravel and I was surprised on how good setting the controllers as RESTful is, it made routes and my code cleaner. I'm kinda new on web development and never used REST before since all my clients' projects are basically CRUD operations. There's any cool buzzword to this "approach" or I'm just stupid for doing it? I don't plan to follow any REST patterns, just to make my life easier and code cleaner. Basicallly just GET/POST, the other ones are not native anyway so (emulated on hidden form value).

  Read the article

• How do I make the jump from Android to Windows Phone 7?

  - by Rob S.

  I'm planning on making the jump over from developing apps for Android to developing apps for Windows Phone 7 as well. For starters, I figured I would port over my simplest app. The code itself isn't much of a problem. The transition from Java to C# isn't that bad. It's actually easier than I expected. What is troublesome is switching SDKs. I've already compiled some basic Windows Phone 7 apps and ran through some tutorials but I'm still feeling a bit lost. For example, I'm not sure what the equivalent of a ScrollView on Android would be on Windows Phone 7. So does anyone have any advice or any resources they can offer me to help me make this transition? Additionally, any comments on the Windows Phone 7 app market (especially in comparison to the Android market) would be greatly appreciated as well. Thank you very much in advance for your time.

  Read the article

• .mdf Database Filetype

  - by James Izzard

  Would somebody be kind enough to correct my understanding of the following (if incorrect)? Microsoft's .mdf file-type can be used by both the LocalDB and the full Server database engines (apologies if engine is not the correct word?). The .mdf file does not care which of these two options are accessing it - so you could use either to access any given .mdf file, provided you had permissions and password etc. The LocalDB and the SQL Server are two options that can be interchangeably chosen to access .mdf files depending on the application requirements. Appreciate any clarification. Thanks

  Read the article

• Software monetization that is not evil

  - by t0x1n

  I have a free open-source project with around 800K downloads to date. I've been contacted by some monetization companies from time to time and turned them down, since I didn't want toolbar malware associated with my software. I was wondering however, is there a non-evil way to monetize software ? Here are the options as I know them: Add a donation button. I don't feel comfortable with that as I really don't need "donations" - I'm paid quite well. Donating users may feel entitled to support etc. (see the second to last bullet) Add ads inside your application. In the web that may be acceptable, but in a desktop program it looks incredibly lame. Charge a small amount for each download. This model works well in the mobile world, but I suspect no one will go for it on the desktop. It doesn't mix well with open source, though I suppose I could charge only for the binaries (most users won't go to the hassle of compiling the sources). People may expect support etc. after having explicitly paid (see next bullet). Make money off a service / community / support associated with the program. This is one route I definitely don't want to take, I don't want any sort of hassle beyond coding. I assure you, the program is top notch (albeit simple) and I'm not aware of any bugs as of yet (there are support forums and blog comments where users may report them). It is also very simple, documented, and discoverable so I do think I have a case for supplying it "as is". Add affiliate suggestions to your installer. If you use a monetization company, you lose control over what they propose. Unless you can establish some sort of strong trust with the company to supply quality suggestions (I sincerely doubt it), I can't have that. Choosing your own affiliate (e.g. directly suggesting Google Toolbar) is possibly the only viable solution to my mind. Problem is, where do I find a solid affiliate that could actually give value to the user rather than infect his computer with crapware? I thought maybe Babylon (not the toolbar of course, I hate toolbars)?

  Read the article

• Library Organization in .NET

  - by Greg Ros

  I've written a .NET bitwise operations library as part of my projects (stuff ranging from get MSB set to some more complicated bitwise transformations) and I mean to release it as free software. I'm a bit confused about a design aspect of the library, though. Many of the methods/transformations in the library come with different endianness. A simple example is a getBitAt method that regards index 0 as the least significant bit, or the most significant bit, depending on the version used. In practice, I've found that using separate functions for different endianness results in much more comprehensible and reusable code than assuming all operations are little-endian or something. I'm really stumped regarding how best to package the library. Should I have methods that have LE and BE versions take an enum parameter in their signature, e.g. Endianness.Little, Endianness.Big? Should I have different static classes with identically named methods? such as MSB.GetBit and LSB.GetBit On a much wider note, is there a standard I could use in cases like this? Some guide? Is my design issue trivial? I have a perfectionist bent, and I sometimes get stuck on tricky design issues like this... Note: I've sort of realized I'm using endianness somewhat colloquially to refer to the order/place value of digital component parts (be they bits, bytes, or words) in a larger whole, in any setting. I'm not talking about machine-level endianness or serial transmission endianness. Just about place-value semantics in general. So there isn't a context of targeting different machines/transmission techniques or something.

  Read the article

• (PHP vs Python vs Perl) vs Ruby [closed]

  - by Dr.Kameleon

  OK, here's what : I've programmed in over 20 different languages and now, because of a large project I'm currently working on for Mac OS X (in Objective-C/Cocoa), I need to make a final decision on which language to use for my background scripting + plugin functionality. Definitely, one factor that'll ultimately influence my decision is which one I'm most familiar with, which is PHP (one of the ugliest languages around, which I however adore... lol), then Python / Perl (the "proven values"... )... and then Ruby (which, to me, is almost confusing and I've only played with it for some time.) Now, here's my considerations : (As previously mentioned) Being familiar with it (anyway, if X is better in my case, I really don't mind studying it from scratch...) Speed Good interaction with the Shell + ease of integration with my Cocoa application Btw, some of the reasons that made me wonder if Ruby would be a good choice is : The hype around it (although, I still don't get why; but that's probably just me...) My major competitor (we're actually talking about the same type of software here) is using Ruby for its backend scripting almost exclusively (ok, along with some BASH). Isn't Ruby considered slower e.g. than Perl? Why did he choose that? Simply, a matter of personal taste? So... your thoughts?

  Read the article

• Rule of thumb for cost vs. savings for code re-use

  - by Styler

  Is it a good rule of thumb to always write code for the intent of re-using it somewhere down the road? Or, depending on the size of the component you are writing, is it better practice to design it for re-use when it makes sense with regards to time spent on it. What is a good rule of thumb for spending extra time on analysis and design on project components that have "some probability" of being needed later down the road for other things that may or may need this part. For example, if I have the need for project X to do things A, and B. A definitely needs to be written for re-use because it just makes sense to do so. B is very project specific at the moment, and I can hack it all together in a couple days to finish the project on time and give everyone kudos for being a great team, etc. Or if we say, lets spend a whole friggin' 2 weeks figuring out what project Y/Z might need this thing for and spend a load of extra time on on part B because someday we might need to use it on project Y/Z (where the savings will be realized). I'd imagine a perfect world situation would be a nicely crafted combination of project specific vs. re-use architected components given the project. However some code shops might feel it would be a great idea to write everything for the intention of using it at some point down the road.

  Read the article

• Grouping a comma separated value on common data [closed]

  - by Ankit

  I have a table with col1 id int, col2 as varchar (comma separated values) and column 3 for assigning group to them. Table looks like col1 col2 group .............................. 1 2,3,4 2 5,6 3 1,2,5 4 7,8 5 11,3 6 22,8 This is only the sample of real data, now I have to assign a group no to them in such a way that output looks like col1 col2 group .............................. 1 2,3,4 1 2 5,6 1 3 1,2,5 1 4 7,8 2 5 11,3 1 6 22,8 2 The logic for assigning group no is that every similar comma separated value of string in col2 have to be same group no as every where in col2 where '2' is there it has to be same group no but the complication is that 2,3,4 are together so they all three int value if found in any where in col2 will be assigned same group. The major part is 2,3,4 and 1,2,5 both in col2 have 2 so all int 1,2,3,4,5 have to assign same group no. Tried store procedure with match against on col2 but not getting desired result Most imp (I can't use normalization, because I can't afford to make new table from my original table which have millions of records), even normalization is not helpful in my context. This question is also on stackoverflow with bounty on this link Achieved so far:- I have set the group column auto increment and then wrote this procedure:- BEGIN declare cil1_new,col2_new,group_new int; declare done tinyint default 0; declare group_new varchar(100); declare cur1 cursor for select col1,col2,`group` from company ; DECLARE CONTINUE HANDLER FOR NOT FOUND SET done=1; open cur1; REPEAT fetch cur1 into col1_new,col2_new,group_new; update company set group=group_new where match(col2) against(concat("'",col2_new,"'")); until done end repeat; close cur1; select * from company; END This procedure is working, no syntax mistake but the problem is that I am not achieving the desired result exactly.

  Read the article

• Algorithm to Solve Most of a Problem

  - by Mike G

  I need an Algorithm/Design Pattern that allows me to try to get the maximum number of rules followed. So I have a couple teams and I need to pair them with a referee and against each other into a round robin. There a rules on who can compete with who and who can judge who so I need to find the configuration that satisfies the most of these. Some rules are more important than others and are "worth more" when evaluating "what satisfies the most of them" There probably isn't a algorithm for this, but is there a design pattern that could help me maximize my chances of finding this configuration?

  Read the article

• Application qos involving priority and bandwidth

  - by Steve Peng

  Our manager wants us to do applicaiton qos which is quite different from the well-known system qos. We have many services of three types, they have priorites, the manager wants to suspend low priority services requests when there are not enough bandwidth for high priority services. But if the high priority services requests decrease, the bandwidth for low priority services should increase and low priority service requests are allowed again. There should be an algorithm involving priority and bandwidth. I don't know how to design the algorithm, is there any example on the internet? Somebody can give suggestion? Thanks. UPDATE All these services are within a same process. We are setting the maximum bandwidth for the three types of services via ports of services via TC (TC is the linux qos tool whose name means traffic control).

  Read the article

• generate parent child relation from the array to print a multi-level menu?

  - by Karthick Selvam

  How to get parent child relation from this array to print a multi-level menu $menus = array ( 0 => array ( 'id'=>0, 'check' => 1, 'display' =>'Arete Home', 'ordering' => -10, 'parent' => none, ), 1 => array ( 'id'=>1, 'check' => 1, 'display' => 'Submit Paper', 'ordering' => -10, 'parent' => 2, 'subordering' => -10, ), 2 => array ( 'id'=>2, 'check' => 1, 'display' => 'Buy Now', 'ordering' => -10, 'parent' => 1, 'subordering' => -10, ), 1461 => array ( 'id'=>1461, 'check' => 1, 'display' => 'Where are We?', 'ordering' => -10, 'parent' => 2, 'subordering' => -10, ), 1463 => array ( 'id'=>1463, 'check' => 1, 'display' =>' About Me?', 'ordering' => -10, 'parent' => 2, 'subordering' => -10, ), 1464 => array ( 'id'=>1464, 'check' => 1, 'display' => 'About You?', 'ordering' => -10, 'parent' => 2, 'subordering' => -10, ), 1465 => array ( 'id'=>1465, 'check' => 1, 'display' => 'About who?', 'ordering' => -10, 'parent' => 1, 'subordering' => -10, ), ); code sample: foreach($menus as $id=>$values) { $values['parent']=isset($values['parent']) ? $values['parent'] : 0; $menus[$values['parent']]['childs'][$id]=$values; unset($menus[$id]); } foreach($menus as $id1=>$value2) { $value2['parent']=isset($value2['parent']) ? $value2['parent'] : 0; $menus[$value2['parent']]['childs'][$id1]=$value2; unset($menus[$id1]); }

  Read the article

• Guiding Management to the Correct Decision

  - by Blumer

  My supervisor (also a developer) and I have a running joke about writing a book called "Managing From Beneath: Subversively Guiding Management to the Right Decision" and including a number of "techniques" we've developed for helping those who make the decisions to make the right ones. So far, we've got (cynicism warning!): BIC It! BIC stands for "Bury In Committee." When a bad idea comes up that someone wants to champion, we try to get it deferred to a committee for input. Typically it will either get killed outright (especially if other members of the committee are competing for you as a resource), or it will be hung up long enough that the proponent forgets about it. Smart, Stupid, or Expensive? When someone gets a visionary idea, offer them three ways to do it: a smart way, a stupid way, and an expensive way. The hope is that you've at least got a 2/3 shot of not having to do it the way that makes a piece of your soul die. All-Pro. It's a preemptive pro/con list in which you get into the mind of the (pr)opponent and think what would be cons against doing it your way. Twist them into pros and present them in your pro list before they have a chance to present them as cons. Dependicitis. Link pending decisions together, ideally with the proponent's pet project as the final link in the chain. Use this leverage to force action on those that have been put off. Preemptive Acceptance. Sometimes it's clear that management is going to go a particular direction regardless of advice to the contrary, and it's time to make the best of it. Take the opportunity to get something else you need, though. Approach the sponsor out of the blue and take the first step: "You know, I've been thinking about it, and while it's not the route I would advise, as long as we can get the schedule and budget for Project Awesome loosened up, I can work some magic to make your project fly." So ... what techniques have you come up with to try to head off the problem projects or make the best of what may come?

  Read the article

• Why is programming sometimes viewed as a second-rate role?

  - by CaptainCodeman

  I've been a programmer for most of my life. I recently interviewed for a management job in a company and the interviewer looked at my CV asked me "How do we know you're not just a programmer". Which in my opinion is quite a rude thing to say, but it's not an isolated incident and I've heard other similar things in other settings. It does seem that for some reason being a programmer is viewed as having a lower station, especially in settings where they have a separate IT department which is viewed as a support role. Is a career in software development doomed to being a second-rate support citizen?

  Read the article

• Session serialization in JavaEE environment

  - by Ionut

  Please consider the following scenario: We are working on a JavaEE project for which the scalability starts to become an issue. Up until now, we were able to scale up but this is no longer an option. Therefore we need to consider scaling out and preparing the App for a clustered environment. Our main concern right now is serializing the user sessions. Sadly, we did not consider from the beginning the issue and we are encountering the following excetion: java.io.WriteAbortedException: writing aborted; java.io.NotSerializableException: org.apache.catalina.session.StandardSessionFacade I did some research and this exception is thrown because there are objects stored on the session which does not implement the Serializable interface. Considering that all over the app there are quite a few custom objects which are stored on the session without implementing this interface, it would require a lot of tedious work and dedication to fix all these classes declaration. We will fix all this declarations but the main concern is that, in the future, there may be a developer which will add a non Serializable object on the session and break the session serialization & replication over multiple nodes. As a quick overview of the project, we are developing using a home grown framework based on Struts 1 with the Servlet 3.0 API. This means that at this point, we are using the standard session.getAttribute() and session.setAttribute() to work with the session and the session handling is scattered all over the code base. Besides updating the classes of the objects stored on session and making sure that they implement the Serializable interface, what other measures of precaution should we take in order to ensure a reliable Session replication capability on the Application layer? I know it is a little bit late to consider this but what would be the best practice in this case? Furthermore, are there any other issues we should consider regarding this transition? Thank you in advance!

  Read the article

• Interconnect nodes in a Java distributed infrastructure for tweet processing

  - by David Moreno García

  I'm working in a new version of an old project that I used to download and process user statuses from Twitter. The main problem of that project was its infrastructure. I used multiple instances of a java application (trackers) to download from Twitter given an specific task (basically terms to search for), connected with a central node (a web application) that had to process all tweets once per day and generate a new task for each trackers once each 15 minutes. The central node also had to monitor all trackers and enable/disable them under user petition. This, as I said, was too slow because I had multiple bottlenecks, so in this new version I want to improve the infrastructure and isolate all functionalities in specific nodes. I also need a good notification system to receive notifications for any node. So, in the next diagram I show the components that I'll need in this new version: As you can see, there are more nodes. Here are some notes about them: Dashboard: Controls trackers statuses and send a single task to each of them (under user request). The trackers will use this task until replaced with a new one (if done, not each 15 minutes like before). Search engine: I need to store all the tweets. They are firstly stored in a local database for each tracker but after that I'm thinking on using something like Elasticsearch to be able to do fast searches. Tweet processor: Just and isolated component with its own database (maybe something like the search engine to have fast access to info generated by the module). In the future more could be added. Application UI: A web application with a shared database with the Dashboard (mainly to store users information and preferences). Indeed, both could be merged into a single web. The main difference with the previous version of the project is that now they will be isolated and they will only show information and send requests. I will not do any heavy task in them (like process tweets as I did before). So, having this components, my main headache is how to structure all to not have to rewrite a lot of code every time I need to access any new data. Another headache is how can I interconnect nodes. I could use sockets but that is a pain in the ass. Maybe a REST layer? And finally, if all the nodes are isolated, how could I generate notifications for each user which info is only in the database used by the Application UI? I'm programming this using Java and Spring (at least I used them in the last version) but I have no problems with changing the language if I can take advantage of a tool/library/engine to make my life easier and have a better platform. Any comment will be appreciated.

  Read the article

• What is the precise definition of programming paradigm?

  - by Kazark

  Wikipedia defines programming paradigm thus: a fundamental style of computer programming which is echoed in the descriptive text of the paradigms tag on this site. I find this a disappointing definition. Anyone who knows the words programming and paradigm could do about that well without knowing anything else about it. There are many styles of computer programming at many level of abstraction; within any given programming paradigm, multiple styles are possible. For example, Bob Martin says in Clean Code (13), Consider this book a description of the Object Mentor School of Clean Code. The techniques and teachings within are the way that we practice our art. We are willing to claim that if you follow these teachings, you will enjoy the benefits that we have enjoyed, and you will learn to write code that is clean and professional. But don't make the mistake of thinking that we are somehow "right" in any absolute sense. Thus Bob Martin is not claiming to have the correct style of Object-Oriented programming, even though he, if anyone, might have some claim to doing so. But even within his school of programming, we might have different styles of formatting the code (K&R, etc). There are many styles of programming at many levels. So how can we define programming paradigm rigorously, to distinguish it from other categories of programming styles? Fundamental is somewhat helpful, but not specific. How can we define the phrase in a way that will communicate more than the separate meanings of each of the two words—in other words, how can we define it in a way that will provide additional meaning for someone who speaks English but isn't familiar with a variety of paradigms?

  Read the article

< Previous Page | 259 260 261 262 263 264 265 266 267 268 269 270  | Next Page >