You break into company X You insert a dns record into their DNS records, eg barclays bank. All hits on Barclays Bank then go to the IP address coded in the local DNS.
THe effects are localised but if you broke into a large ISP and did this it would catch a lot of people.
This is a real concern. The whole process of programming 'should' be egoless. The person who writes and maintains the code often feels to much ownership and instead of taking the the change in good grace and thankfulness they often approach the whole thing defensively. "The code was never intended to do that".
This is NOT a problem with Open Source development but with programmers as a whole, myself included but I try and suppress it. You have to 'give up' code that you have too much ownership in.
The new employer will always wonder whether you quit before you were fired. There is always the psychology of the employer 'stealing' an employee from another company.
Always, always get your next job first unless you are a contractor then people accept market forces dictates a holiday sometimes.
If you become contractor don't expect taking full time job again will be easy. Employers think that you are filling in a gap between contracts so wont hire you.
Looks like a redundant question since it reads like you have already resigned. What would I do now in your position. Take any job I can that looks in the right direction. If you are as good as you think you are the rewards will come and any job good or bad is experience and will at least educate you in how not to do things in the future.
Mathematically compliment is opposite sides of an angle. if one is acute then the other must be obtuse. They compliment each other like the chinese ying and yang.
They balance one another because they are different or compliment each other. THis is different from the compliment that sings the praise of someone else.
Theneat question is how do you know when to drill into the code and when to fly at a high level.
I regularly work on large suites and I work successfully because I don't step through in detail line by line. I overview the lot get a general structure then target where I think problems are more likely.
In Linux case there are probably clues to what the problem is, only with this driver, only with this option, etc. I have the utmost respect for any kernel developer.
BOth coding styles are nessecary in real life. I often put in a patch to keep things going then rework the code later under another change to make the code clean again.
Refactoring poor solutions is important but in a stable environment you do not want a chunk of code rewritten or even just reformatted although the code is poorly indented. You want small easily review chunks of code that are obvious what it is fixing.
Firstly USL, Novell, gave the Unix trademark away before selling the distribution rights of the Unix Software. So no SCO does not own the trademark.
Linux is not Unix. This is a repeating theme from a number of people. It is a look alike system not derived from Unix and hence not 'Unix'. This distinction may not be relevant to the programmer but it is a very important legal point.
The program that you write in an interpretter is not executed in the classic sense of creating actual binary. It is more a case of looking at the data portion then doing something in the execution space. Your 'program' remains in data, whether in text or some intermediate compiled form.
As noted this is NOT related to the vendors, RedHat or Microsoft for example, it is related to the vendors of the hardware. As Linux becomes more mainstream, and I think everyone would agree that this is occurring, then the vendors will start supporting Linux more directly eliminating that problem entirely.
I strongly recommend that you try some of the latest live CD's on your hardware. I have been impressed with the improvements in hardware support over the last 12 months so your comments may be dated.
I never read the original article however the response was a work of art in it's own right and very genuinely funny. This response was a highlight of my weekend and can be treated as a standalone piece of real geek humour.
I have exactly the reverse experience. I have a centrino laptop and it performs poorly where Linux says the signal is good and pulls downloads very quickly. In my bedroom opposite ends of the house with bricks walls and steel between Windows XP is extremely flakey with things cancelling out all the time. Linux just goes slowly but does not stop.
Kernel 2.6.10 from Ubuntu. I used to compile my own but now I just don't bother. I even have a pretty GUI to set up the wireless networking options.
If you want to wireless you need good quality cards. If you are buying cards without reading reviews you are doomed to failure. I have a dlink base that I just don't use it wont work, my linksys worked straight out of the box.
Excel 4.0 exporting to csv took the format and exported #### displayed on the screen instead of the underlying numbers and lost me the data. I had to rebuild from scratch.
Contacted MS they were very sorry, we note the problem and will give you a free upgrade to 5.0 when it comes out. Still waiting... I have no idea if the bug still exists.
IE recently V5. If you include two.JS files into a web page it locks up hard. I wrote and debugged the pace in mozilla because of the better debugger and it took me ages to figure out the real problem.
Contacted MS after figuring out the problem and gave a detailed description of the problem and the fact that I no longer needed to solve the problem. In order to submit a detailed bug report I had to pay a fee ($135 springs to mind) and MS would refund it if they agreed that it was their problem. I politely declined and argued over 4 emails that I no longer had a problem however it shoudl be a documented bug in MSDN and preferably it should be fixed so that others did not have to search for the problem. Net result, I have no idea whether my bug report went to the developers nor whether it will be fixed. I kind of doubt it.
I agree look at how they (don't) stand by their EULA.
Lets compare...
Borland removed their bug reporting software off the web entirely. Last time I checked I could not figure out how to report a bug to them.
gcc patched 1 problem overnight, 24 hours. second problem took three days from report to fix.
Reporting a useabilty bug on evolution. Hey my bug has been there a while, pretty minor problem but any developer can find the bug so one day it will be fixed.
I wrote an application that worked on Windows, Unix, Linux, and mainframe. I wrote the GUI separately calling the library function then a simple command line version for Unix.
What I should have done is totally separate the GUI and the application entirely. I could rewrite the GUI in in a portable framework on linux faster than any port of the Windows software. I could then install the framework on all the windows desktops and have them work the same. Not port but total rewrite in a portable framework.
Problem with a lot of GUI code is that the business logic becomes intertwined with the GUI logic. It is incredibly easy to do especially with the Rapid Application Development (RAD) tools and this cause nightmares in porting if your RAD tool is not cross platform.
One of the problem with ALL source control systems is the lack of portable change sets.
I would love to be able to duplicate a project and then import their changesets into mine keeping it synced no matter what source control they are using. This would save bandwidth because I only need changes not the whole repository.
To my knowledge there is NO standard for change sets and if there are it certainly is not actively being used.
Linux has come a long way in terms of packaging and dependencies. It may be able to be installed easily because the legacy packages are avaialble or it may take a little research however I could find those old libraries and run it.
More of a problem with Windows is when the dll is called the same this with a different binary interface. This happens and causes a lot of those crashes that you see. Linux on the other hand has a model that shows when a shared library (equivalent to a dll) has a binary incompatible upgrade and two versions can happily live on the same machine servicing both programs.
Finally most applications can be recompiled or drawn from current packages getting rid of bugs and security problems for free.
So all in all I would say that the fundamental process of Linux is far more robust from a technical viewpoint.
I realised the problem with this article is that it is actually wrong.
We want to use off the shelf solutions that are best for our business (or person) whether they are commercial or not. I believe that CPAN is the best off the shelf solutions available today in a clear easily searchable fashion. I want to use this because it is good software (mostly) and it works. It is not commercial so it is not COTS.
Kernels of major distros have all the required drivers built as modules so compiling is really a thing of the past. Sure there are a few exceptions but for comercial applications they tend not to have bleeding edge systems very often and when they do it is purpose ordered so it must work out of the box with a linux distro.
So your issues with hardware are not 100% typical. I bought a laptop and installed ubuntu linux on it and it 'just worked' without any tweaking.
I put up with it because it does work and is secure.
I run Solaris, AIX, Linux, cygwin and standard windows CVS.
If you have multiple platforms and multiple copies sitting on those separate machines then a source repository is a must. I spent a huge amount of time merging my changes across versions on separate platforms now I simply cvs up and it all works, well I do make portability mistakes but this is easier to fix.
Backing up your cvs server is simply a matter of backing up the directories that is your repository. You should NEVER have to restore it, you can get an old version of the code, and overwrite the current version to revert something. This gives you a trail of failures as well as sucessful code changes.
If you use cvs over ssh then you probably can leave it a LONG time between patches. Instead of learning cvs I would suggest looking at other more developed tools like ARCH and subversion which have more advanced capabilities first. They may provide something that you need.
I beleive that everyone who is excellent is self taught. Why because they do not stop when they leave school. If you stagnate then it does not matter if you are self taught or a college major.
Did they do 'just enough' to get something working and keep their job or did they continue to learn. I have been self taught for 25 years. Every day I learn something new. Test driven development, testing principles, programming, even inspections of code.
Ohhh yes, I teach at University part time now.:-) You probably learnt from self taught people.
I have studied a lot over the years and I have come to the opinion that most geeks cannot get egoless programming. Programming is owned by the people you are working for and with (FOSS) not you. This is across the board, self taught and college graduates 'own' code and hate others helping or changing 'their code'. College graduates are typically WORSE because they are taught that sharing is cheating which is a totally wrong premise in the real world. Give me a self taught anyday, this includes Uni students who worked outside the square.
I am a geek. I was having a problem meeting real friends not work aquaintences.
I joined scouts as a scout leader and now I have friends from many walks of life very different to my own but just as interesting and dedicated as I hope I am.
I actually went and met those people on the other side of IRC. It was fabulous sitting in a pub in Ireland with people I had known fro three years, and then the same in Germany. I count these people as friends because they are real and can come and stay at my house anytime.
I was working 16 hours a day 5-6 days a week. Work has NEVER been so happy with me since I started saying no and working 9-5 (well mostly). Working hard does not always mean working better.
Firstly the anouncement is purely for the Version 2.0 codeline. This is an excellent idea because it focusses everyones attention on getting the best Mac Port possible in the timeframe, not scattering resources trying many things.
The Mac effort is one of the most intense efforts in OOo today by FOSS developers. There are many volunteers and almost daily offers for additional help. So as they say, news of my (OOo) death is premature.
Ultimately the NEO office port will be merged with the mainline OOo. At this stage there are some issues with doing this cleanly so it is managed (extremely well) by a third party. This will continue until the whole thing becomes clean enough to merge. Try NEO if that works for you that is still a win for OOo in my book, I do not care about the brand name frankly my effort in making OOo better in a number of small ways is paying off, I am proud.
Finally do not forget that this is an Open Source development. Any predictions that something will not happen are just very unlikely because someone with a bee in his or her bonnet will do what you do not expect. If you want an Aqua port more you want a serious stable Office Suite using X on Mac then please by all means, do that.
Slashdot Mirror
From my reading of the article...
You break into company X
You insert a dns record into their DNS records, eg barclays bank.
All hits on Barclays Bank then go to the IP address coded in the local DNS.
THe effects are localised but if you broke into a large ISP and did this it would catch a lot of people.
This is a real concern. The whole process of programming 'should' be egoless. The person who writes and maintains the code often feels to much ownership and instead of taking the the change in good grace and thankfulness they often approach the whole thing defensively. "The code was never intended to do that".
This is NOT a problem with Open Source development but with programmers as a whole, myself included but I try and suppress it. You have to 'give up' code that you have too much ownership in.
It is always easier to get a job with a job.
The new employer will always wonder whether you quit before you were fired. There is always the psychology of the employer 'stealing' an employee from another company.
Always, always get your next job first unless you are a contractor then people accept market forces dictates a holiday sometimes.
If you become contractor don't expect taking full time job again will be easy. Employers think that you are filling in a gap between contracts so wont hire you.
Looks like a redundant question since it reads like you have already resigned. What would I do now in your position. Take any job I can that looks in the right direction. If you are as good as you think you are the rewards will come and any job good or bad is experience and will at least educate you in how not to do things in the future.
Mathematically compliment is opposite sides of an angle. if one is acute then the other must be obtuse. They compliment each other like the chinese ying and yang.
They balance one another because they are different or compliment each other. THis is different from the compliment that sings the praise of someone else.
Theneat question is how do you know when to drill into the code and when to fly at a high level.
I regularly work on large suites and I work successfully because I don't step through in detail line by line. I overview the lot get a general structure then target where I think problems are more likely.
In Linux case there are probably clues to what the problem is, only with this driver, only with this option, etc. I have the utmost respect for any kernel developer.
BOth coding styles are nessecary in real life. I often put in a patch to keep things going then rework the code later under another change to make the code clean again.
Refactoring poor solutions is important but in a stable environment you do not want a chunk of code rewritten or even just reformatted although the code is poorly indented. You want small easily review chunks of code that are obvious what it is fixing.
Horses for courses as the saying goes.
Firstly USL, Novell, gave the Unix trademark away before selling the distribution rights of the Unix Software. So no SCO does not own the trademark.
Linux is not Unix. This is a repeating theme from a number of people. It is a look alike system not derived from Unix and hence not 'Unix'. This distinction may not be relevant to the programmer but it is a very important legal point.
The program that you write in an interpretter is not executed in the classic sense of creating actual binary. It is more a case of looking at the data portion then doing something in the execution space. Your 'program' remains in data, whether in text or some intermediate compiled form.
I was making a counter point to your statement.
As noted this is NOT related to the vendors, RedHat or Microsoft for example, it is related to the vendors of the hardware. As Linux becomes more mainstream, and I think everyone would agree that this is occurring, then the vendors will start supporting Linux more directly eliminating that problem entirely.
I strongly recommend that you try some of the latest live CD's on your hardware. I have been impressed with the improvements in hardware support over the last 12 months so your comments may be dated.
I never read the original article however the response was a work of art in it's own right and very genuinely funny. This response was a highlight of my weekend and can be treated as a standalone piece of real geek humour.
I have exactly the reverse experience. I have a centrino laptop and it performs poorly where Linux says the signal is good and pulls downloads very quickly. In my bedroom opposite ends of the house with bricks walls and steel between Windows XP is extremely flakey with things cancelling out all the time. Linux just goes slowly but does not stop.
Kernel 2.6.10 from Ubuntu. I used to compile my own but now I just don't bother. I even have a pretty GUI to set up the wireless networking options.
If you want to wireless you need good quality cards. If you are buying cards without reading reviews you are doomed to failure. I have a dlink base that I just don't use it wont work, my linksys worked straight out of the box.
Sounded like a good idea at the time is now a major problem.
Don't bring them back!
You want a track record...
.JS files into a web page it locks up hard. I wrote and debugged the pace in mozilla because of the better debugger and it took me ages to figure out the real problem.
Excel 4.0 exporting to csv took the format and exported #### displayed on the screen instead of the underlying numbers and lost me the data. I had to rebuild from scratch.
Contacted MS they were very sorry, we note the problem and will give you a free upgrade to 5.0 when it comes out. Still waiting... I have no idea if the bug still exists.
IE recently V5. If you include two
Contacted MS after figuring out the problem and gave a detailed description of the problem and the fact that I no longer needed to solve the problem. In order to submit a detailed bug report I had to pay a fee ($135 springs to mind) and MS would refund it if they agreed that it was their problem. I politely declined and argued over 4 emails that I no longer had a problem however it shoudl be a documented bug in MSDN and preferably it should be fixed so that others did not have to search for the problem. Net result, I have no idea whether my bug report went to the developers nor whether it will be fixed. I kind of doubt it.
I agree look at how they (don't) stand by their EULA.
Lets compare...
Borland removed their bug reporting software off the web entirely. Last time I checked I could not figure out how to report a bug to them.
gcc patched 1 problem overnight, 24 hours. second problem took three days from report to fix.
Reporting a useabilty bug on evolution. Hey my bug has been there a while, pretty minor problem but any developer can find the bug so one day it will be fixed.
I wrote an application that worked on Windows, Unix, Linux, and mainframe. I wrote the GUI separately calling the library function then a simple command line version for Unix.
What I should have done is totally separate the GUI and the application entirely. I could rewrite the GUI in in a portable framework on linux faster than any port of the Windows software. I could then install the framework on all the windows desktops and have them work the same. Not port but total rewrite in a portable framework.
Problem with a lot of GUI code is that the business logic becomes intertwined with the GUI logic. It is incredibly easy to do especially with the Rapid Application Development (RAD) tools and this cause nightmares in porting if your RAD tool is not cross platform.
Go to the library and ask them to order the book in. In Australia this is a free service.
They actually can do this, I have used this to read an obscure book I could not get any other way.
One of the problem with ALL source control systems is the lack of portable change sets.
I would love to be able to duplicate a project and then import their changesets into mine keeping it synced no matter what source control they are using. This would save bandwidth because I only need changes not the whole repository.
To my knowledge there is NO standard for change sets and if there are it certainly is not actively being used.
Linux has come a long way in terms of packaging and dependencies. It may be able to be installed easily because the legacy packages are avaialble or it may take a little research however I could find those old libraries and run it.
More of a problem with Windows is when the dll is called the same this with a different binary interface. This happens and causes a lot of those crashes that you see. Linux on the other hand has a model that shows when a shared library (equivalent to a dll) has a binary incompatible upgrade and two versions can happily live on the same machine servicing both programs.
Finally most applications can be recompiled or drawn from current packages getting rid of bugs and security problems for free.
So all in all I would say that the fundamental process of Linux is far more robust from a technical viewpoint.
There are cobol programmers in open source and recent cobol compilers have socket connections. So this is not that difficult.
I realised the problem with this article is that it is actually wrong.
We want to use off the shelf solutions that are best for our business (or person) whether they are commercial or not. I believe that CPAN is the best off the shelf solutions available today in a clear easily searchable fashion. I want to use this because it is good software (mostly) and it works. It is not commercial so it is not COTS.
Declare war on the COTS phrase, it it OTS!
Kernels of major distros have all the required drivers built as modules so compiling is really a thing of the past. Sure there are a few exceptions but for comercial applications they tend not to have bleeding edge systems very often and when they do it is purpose ordered so it must work out of the box with a linux distro.
So your issues with hardware are not 100% typical. I bought a laptop and installed ubuntu linux on it and it 'just worked' without any tweaking.
I put up with it because it does work and is secure.
I run Solaris, AIX, Linux, cygwin and standard windows CVS.
If you have multiple platforms and multiple copies sitting on those separate machines then a source repository is a must. I spent a huge amount of time merging my changes across versions on separate platforms now I simply cvs up and it all works, well I do make portability mistakes but this is easier to fix.
Backing up your cvs server is simply a matter of backing up the directories that is your repository. You should NEVER have to restore it, you can get an old version of the code, and overwrite the current version to revert something. This gives you a trail of failures as well as sucessful code changes.
If you use cvs over ssh then you probably can leave it a LONG time between patches. Instead of learning cvs I would suggest looking at other more developed tools like ARCH and subversion which have more advanced capabilities first. They may provide something that you need.
I beleive that everyone who is excellent is self taught. Why because they do not stop when they leave school. If you stagnate then it does not matter if you are self taught or a college major.
:-) You probably learnt from self taught people.
Did they do 'just enough' to get something working and keep their job or did they continue to learn. I have been self taught for 25 years. Every day I learn something new. Test driven development, testing principles, programming, even inspections of code.
Ohhh yes, I teach at University part time now.
This is pure BS.
I have studied a lot over the years and I have come to the opinion that most geeks cannot get egoless programming. Programming is owned by the people you are working for and with (FOSS) not you. This is across the board, self taught and college graduates 'own' code and hate others helping or changing 'their code'. College graduates are typically WORSE because they are taught that sharing is cheating which is a totally wrong premise in the real world. Give me a self taught anyday, this includes Uni students who worked outside the square.
Self taught and proud of it!
I am a geek. I was having a problem meeting real friends not work aquaintences.
I joined scouts as a scout leader and now I have friends from many walks of life very different to my own but just as interesting and dedicated as I hope I am.
I actually went and met those people on the other side of IRC. It was fabulous sitting in a pub in Ireland with people I had known fro three years, and then the same in Germany. I count these people as friends because they are real and can come and stay at my house anytime.
I was working 16 hours a day 5-6 days a week. Work has NEVER been so happy with me since I started saying no and working 9-5 (well mostly). Working hard does not always mean working better.
Firstly the anouncement is purely for the Version 2.0 codeline. This is an excellent idea because it focusses everyones attention on getting the best Mac Port possible in the timeframe, not scattering resources trying many things.
The Mac effort is one of the most intense efforts in OOo today by FOSS developers. There are many volunteers and almost daily offers for additional help. So as they say, news of my (OOo) death is premature.
Ultimately the NEO office port will be merged with the mainline OOo. At this stage there are some issues with doing this cleanly so it is managed (extremely well) by a third party. This will continue until the whole thing becomes clean enough to merge. Try NEO if that works for you that is still a win for OOo in my book, I do not care about the brand name frankly my effort in making OOo better in a number of small ways is paying off, I am proud.
Finally do not forget that this is an Open Source development. Any predictions that something will not happen are just very unlikely because someone with a bee in his or her bonnet will do what you do not expect. If you want an Aqua port more you want a serious stable Office Suite using X on Mac then please by all means, do that.