I agree. Some of these are very difficult to avoid and people think they are doing you a favour by downloading it. The one that entertained me was telling me my Windows was too slow and that was on a Linux box before I had locked down Mozilla completely.
I live in Germany at the moment and we have some major problems with dialers using high-toll prefixes.
It used to be just porn-sites getting up to the dodgy stuff. Now it seems to be anything (although not on/. yet).
In the long dream sequence of running through a deserted Times Square, the ads were replaced with ones that were more acceptable to the company behind the film.
Ad substitution technology is quite advanced now even for real-time stuff so that TV companies showing sports matches have to be contracturally forbidden from replacing the ads.
I wouldn't worry about GATOR/GAIN's ad pop-over system if they didn't screw up existing software. Maybe they are better now, but if their programmer's aren't deliberately sabotaging systems, then they are just incompetent, assuming their stuff will always work with a particular version of IE.
Gator/GAIN do not advertise that they may interfere with the correct operation of your computer. I have had some real fun experiences removing these programs. Than heavens for ad-ware removal tools.
Reading source code and navigating your way around a large system isn't easy. I work with large and complicated propietary systems as well as open source code. I can't say that one is better documented than the other. Both are equally a problem when getting into the code.
To be honest, from what I can see of this book, it doesn't seem to help that much.
Easily getting into a large project means more than just doing a make tags. Personally, I find an elising editor is great for getting the overview, but I haven't found a good one yet that is open source.
The guy who built the Accra gateway with me, Peter Kennedy, later took a contract job building telecom infrastructure in Chechnya, was taken hostage by Chechen rebels for ransom, and was found decapitated a few weeks later.
My wife worked for the company (Linya Svayzi) in St Petersburg that comissioned the project in Chechnya with the engineers supplied by Granger Telecom. It was installing a GSM system, if I remember right. Infrastructure projects in areas like Chechnya were seen as cooperating with the regime and engineers were definitely targets by those who were against the regime.
The word is that the amount originally asked for wasn't big (about $5000) but it was nixed by the Russian government who didn't want any ransoms paid to terrorists.
To be fair the US State dept had nothing to do with this one as the engineers were British and Kiwi. The UK govt had already issued warnings about visiting Chechnya. Liniya Svayzi definitely knew that Chechny was dodgy. Granger had worked in Russia before so they must have been aware of the relative risks.
This year there are some stages in the Pyranees, the Alpes-Maritimes and the Alps. On the eighth stage this year, they will cross a pass at 2650m. Not unusual.
I don't see it being very easy either going up that mountain or down it on a recumbent. Racing for endurance with gentle hills may be somthing else, but otherwise the traditional bike seems somewhat better.
Is compact flash self correcting? I would have though that the incidental radiation would not be too nice.
Otoh, non-rotating disk memory is fast and only uses a disk during startup and shutdown (whether failure or otherwise). Normally such solid state disks have ECC and so on and they are much faster than conventional disks. You can also RAID them for additional reliability.
Good point, but you could send back a "service not available" ISDN code if you have a software call switch. As part of the incoming call negotiation, the caller provides number and service required. If you send back the N/A message (you get this when trying to call an ISDN fax with a voice phone) then the call is bounced by the telco.
Abstracting things is a real nice thing to do. Encapsulating is also nice and when we get to LRPCs hooking code between processes, it is^really cute.
The trouble is that by this time you are going through so many virtual function tables, process switches or whatver to do something basic you are completely screwed over for performance. Oh and given all the DLLs that need dynamic linking with your image when you activate - what chance do you get?
The kernel of Win2K/NT is quite nice. If someone sat down and coded a Linux style API on top of it, it could be serious competition. They tried to do this with their Posix layer. However, they screwed it up in other ways.
If anyone ever gets to meet Dave Cutler, the primary NT architect, just ask him how he feels about performance and memory use now against the days when he wrote an operating system for the PDP-11 (16-bit, 64K and small).
GnuCash has basic business methods coded in a very C++ish dialect of C. However the main stuff is coded in Scheme making it relatively easy to extend. It can be a little slow to load though. It also has a bit done in Perl.
Congratulations on a successful project. However, you mentioned what worked without commenting on it:
Two years ago we took the descision to re-design the toolkit from the ground up with as much input from as many people as possible. Since that time we have strived to make sure that as many people as possible have an input into the design process and we keep that process as open as possible by pubishing the IRC sessions in which discussions take place.
The moment you involve other people in a project, you need to document and explain design decisions and any discussions arising. By doing this you have made it very easy for other developers to get involved. I also like the way that you publish IRC discussions.
You are almost a text book example on how to do things right.
Visual Basic isn't bad for RAD, particularly on clients and it has a great IDE/Design tool. But it is MS only. In the end, the open source alternatives are better if only because you have a reasonable expection of being platform agnostic. Other RAD languages such as Delphi, really have the same problem.
Personally, I feel the key point is extendability. In the end you have some real program code doing the 10% of an ap that is performance critical. That is probably in C or C++. It is easy to call out to this in Perl/Python/Ruby or whatever. Also what Perl in particular has, is CPAN. Some algorithms may be suboptimal or non-portable (not everything there runs on Windows), but it is axcellent starting point.
We don't really need VB on Linux, what we need is for someone to come up with a better IDE for what we have.
Exactly. If you as a developer make use of LGPL software, and you then incorporate that into a commercial package, you assume liability for it. This is reasonable and not new.
If I buy a RedHat or Suse distribution, everything is freely downloadable anyway. What I pay for is the 'packaging' and support.
As for author's rights. Well you can't disown your work but you can reassign the copyright. As for collective rights, well again this is no problem in the book publishing field. Why should it be for software?
Personally, I just see this as another attempt to spread FUD.
I am not a German lawyer but have some idea of my rights as I live there. Under German and EU law, the *vendor* is liable for problems in any software sold. The vendor has recorse against the supplier and then against the developer when the software has been paid for. If at some point the software wasn't sold, i.e., an LGPL package, then there is *no* corresponding obligation on the next level of the chain.
The NT/2K kernel probably is. I have seen Cutler's code from when he worked at Digital (source listings were easy to get then) and it was generally extremely good (Don't ask about his PL/1 compiler though). I don't think he would let his team members get away with much less even if he isn't writing it.
However the kernel is just one part and unless you are writing drivers, you never see it as it sits behind the smoke and mirrors of many different (and not necessarily consistent) APIs.
I agree with you about the Troll. It is probably a business wannabe.
You forget to mention though that all the shennigans now happening come down to some unsuccessful closed source companies getting very worried. Some of them produced reasonable software, but they can't support it any more. Support on OSS is variable too, but at least I've got the source code. Some companies using the support model to get revenue continuity from older products really don't want to lose this.
The walls probably aren't really that thick unless you live in a bank vault or a medieval cottage (thin on the ground in the US). The wall is probably just has a cavity. With a long drill bit and taking time, it is still quite possible to drill through such walls with a domestic drill (750w).
Wireless is fine until you want networked file systems.
It sounds dumb, but cable pulling isn't easy without having something to pull it with. First thread string through holes using stiff wire and then tie the string to the cable and pull that through (leaving sufficient string to pull further cables).
Another useful tool for the initial string threading is flexible curtain rail. It is quite stiff and able to be threaded for quite a long way.
I went to high school in the UK, so it is quite possible that they had different ideas about Tolkein. He was, after all a Professor of English at Oxford which gave him a certain amount of 'cred' amongst the teachers along with others of his circle such as C.S. Lewis.
Later fantasy works such as Peake's Gormanghast were also suggested reading material, however fantasy was comparatively underemphasised on the reading list. I hear otherwise now.
seems to ba a major problem. Jo Rowling's success has made many people jealous.
However, any author, publisher or critic who is really concerned about their craft gives her a nod because she is introducing kids to the addictive habit of reading.
With the skewed release schedules for the translated versions, she is also forcing many impatient kids around the world to read the book in a foreign language (thus improving their English).
I live in Germany at the moment and we have some major problems with dialers using high-toll prefixes.
It used to be just porn-sites getting up to the dodgy stuff. Now it seems to be anything (although not on /. yet).
Ad substitution technology is quite advanced now even for real-time stuff so that TV companies showing sports matches have to be contracturally forbidden from replacing the ads.
Gator/GAIN do not advertise that they may interfere with the correct operation of your computer. I have had some real fun experiences removing these programs. Than heavens for ad-ware removal tools.
To be honest, from what I can see of this book, it doesn't seem to help that much.
Easily getting into a large project means more than just doing a make tags. Personally, I find an elising editor is great for getting the overview, but I haven't found a good one yet that is open source.
I don't see it being very easy either going up that mountain or down it on a recumbent. Racing for endurance with gentle hills may be somthing else, but otherwise the traditional bike seems somewhat better.
Thanks, then it looks like it might be in with a chance for 2.6. I am surprised though that it was merged in at at all in that condition.
Is the framebuffer stuff fixed yet? Last time I liked (around 2.5.6x or so) it was very badly broken.
Otoh, non-rotating disk memory is fast and only uses a disk during startup and shutdown (whether failure or otherwise). Normally such solid state disks have ECC and so on and they are much faster than conventional disks. You can also RAID them for additional reliability.
Good point, but you could send back a "service not available" ISDN code if you have a software call switch. As part of the incoming call negotiation, the caller provides number and service required. If you send back the N/A message (you get this when trying to call an ISDN fax with a voice phone) then the call is bounced by the telco.
The trouble is that by this time you are going through so many virtual function tables, process switches or whatver to do something basic you are completely screwed over for performance. Oh and given all the DLLs that need dynamic linking with your image when you activate - what chance do you get?
The kernel of Win2K/NT is quite nice. If someone sat down and coded a Linux style API on top of it, it could be serious competition. They tried to do this with their Posix layer. However, they screwed it up in other ways.
If anyone ever gets to meet Dave Cutler, the primary NT architect, just ask him how he feels about performance and memory use now against the days when he wrote an operating system for the PDP-11 (16-bit, 64K and small).
GnuCash has basic business methods coded in a very C++ish dialect of C. However the main stuff is coded in Scheme making it relatively easy to extend. It can be a little slow to load though. It also has a bit done in Perl.
You are almost a text book example on how to do things right.
Personally, I feel the key point is extendability. In the end you have some real program code doing the 10% of an ap that is performance critical. That is probably in C or C++. It is easy to call out to this in Perl/Python/Ruby or whatever. Also what Perl in particular has, is CPAN. Some algorithms may be suboptimal or non-portable (not everything there runs on Windows), but it is axcellent starting point.
We don't really need VB on Linux, what we need is for someone to come up with a better IDE for what we have.
If I buy a RedHat or Suse distribution, everything is freely downloadable anyway. What I pay for is the 'packaging' and support.
As for author's rights. Well you can't disown your work but you can reassign the copyright. As for collective rights, well again this is no problem in the book publishing field. Why should it be for software?
Personally, I just see this as another attempt to spread FUD.
I am not a German lawyer but have some idea of my rights as I live there. Under German and EU law, the *vendor* is liable for problems in any software sold. The vendor has recorse against the supplier and then against the developer when the software has been paid for. If at some point the software wasn't sold, i.e., an LGPL package, then there is *no* corresponding obligation on the next level of the chain.
However the kernel is just one part and unless you are writing drivers, you never see it as it sits behind the smoke and mirrors of many different (and not necessarily consistent) APIs.
You forget to mention though that all the shennigans now happening come down to some unsuccessful closed source companies getting very worried. Some of them produced reasonable software, but they can't support it any more. Support on OSS is variable too, but at least I've got the source code. Some companies using the support model to get revenue continuity from older products really don't want to lose this.
If this was one off and signed, I would debate with him, but it is just a troll (a little like BSD/xxx is dying). It could even be the same troll.
This is one reason why I really don't like MS Exchange. Everything collapses, but if I have sources, it makes recovery easier.
Wireless is fine until you want networked file systems.
Seriously, I agree, why get a contractor and why write an article about it. I use contractors at work, but every metre of Cat-5 at home is mine.
Another useful tool for the initial string threading is flexible curtain rail. It is quite stiff and able to be threaded for quite a long way.
Later fantasy works such as Peake's Gormanghast were also suggested reading material, however fantasy was comparatively underemphasised on the reading list. I hear otherwise now.
With the skewed release schedules for the translated versions, she is also forcing many impatient kids around the world to read the book in a foreign language (thus improving their English).