For some reason I naively assumed it would become a link automatically. If only I would learn to preview...
Re:Java's not exactly pining for the fields just n
on
Java vs .NET
·
· Score: 3, Interesting
By coincidence, I gave a talk on how Java and.NET compare, just a few days ago:
http://kylecordes.com/story-260-java-net-talk.ht ml
One thing with the SD Times identified as a source of trouble (the size of the included libraries), I identified as a strength:.NET has a lot of things "in the box" that a developer / team needs to go find, select, download, install, etc. for Java. I personally like the best-of-breed approach that can result from that (choose from the many SOAP implementation, choose the best one for our needs), but for many developers, having a good-enough solution "in the box" is more palatable.
This is particularly exacerbated in some development shops where extensive format decision and permission processes are needed to make any third party tool purchase (or free software adoption).
Something important to consider is the noise generated by the fan in a projector. There may be some quiet ones out there, but most projectors I encounted, regardless of DLP/LCD/etc., are too loud. (How loud is too loud? I expect a $2-5000 TV to be at least not much worse than a $200 TV...)
The V1 had a single, large chamber (for a smaller number of larger bangs); the ones in this article have a bunch of small chambers, each running many cycles per second, for combined rate of hundreds of smaller bangs for second. Perhaps, once it's scaled up to production size, all these smaller bangs will blend together in to a roar, which might be similar in loudness to a normal get engine of similar size.
I looked at getting a single behemoth display like that, but decided that two smaller screens (two 19" LCDs) are a much better deal - more pixels for less money.
For some uses, having two screens would be a problem (i.e. graphic editing), but for software development it works quite well.
The essence of the elections is not whether they are conducted with the same set of rules as in the U.S.; the essence is whether there is some semblance of democracy at all.
I would not even expect offhand that Iraq will have a governmental structure or election structure like the U.S. has. Keep in mind that the way the U.S. elections, federal/state system work, etc., are a 200 year old legacy system, not something that the current generation of leaders dreamed up:-)
Somewhere along the way, I believe a NASA person pointed out that there is only a brief time just before landing when bailout is possible. The shuttle comes in very fast, still supersonic even partway across Florida.
Not always; there can be a lot of value in an unenforcable patent, to create a chilling effect on competitors, especially the smaller ones. That's because noone really knows if it's enforcable until someone can afford to spend a hefty sum on litigation to find out.
The way to address this is to hire good people, and treat them well, especially for key positions like this one. That way when the person leaves to move on in their career, it will be with regret and good wishes. They will be eager to help you in way they can help you recover from their departure, rather than looking for ways to harm you.
A while back a resigned from a company. I got a big chuckle out of the HR people handing the "termination". The wording of all the documents was as though they were doing the terminating, which was far from the case (the management and I had and still have a good relationship).
There are of course some similarities between firing someone and them leaving to move on in their career, but treating the latter too much like the former, just makes you look silly.
I have two 191T's in front of me now. They are superb; I haven't noticed any dead pixels yet in either of them, after a couple months of daily use.
I run DVI input on both; for a while I had DVI on one and analog on the other, and the difference was immediately noticable and substantial. Ignore those who claim that DVI in unnecessary; at 1280 resolution, DVI is a large, important, don't-go-without-it difference.
I find 1280 resolution to be quite nice on a desktop 19" LCD. This is probably because I have it positioned at a much farther and more comfortable distance than my 15" laptop display was.
Also 19" LCDs unfortunately cost rather more than $500, for good quality ones anyway.
By using.EXE and.DLL for the various kinds of ready-to-deploy.NET code, Microsoft helped enormously in making it feel native to those who develop and deploy it. You could deploy an app, consisting of an EXE and some DLLs, without even noticing that it's.NET based, particular down the road a bit when the.NET runtime is more widely deployed.
This naming is one of several things that will make.NET feel native, in a way that Java apps do not. It's not a technical difference, it's a cosmetic difference that will make adoption a bit easier.
I'm working under the assumption that their errant software update did not physically damage the cable modem, it just put in some updated software that renders it unusable.
I suggest contacting the manufacturer of the cable modem; there may be a combination of some "reset" step and some kind of firmware update that would overwrite whatever badness the cable company (accidentally, I assume) put on there, and get it going again.
I've recently switched from SpamAssasin to POPFile for the same reason. With SA, once the version I had became slighly outdated, spammers had found ways to write spam that didn't set it off, and a steady flow got through. With POPFile, I generally only ever see one or two of a new flavor of spam, since after a couple of reclassifications the rest gets detected.
Delphi is not very popular for "enterprise" development. However, it is quite popular for shrinkwrap software development, especially vertical market apps. Moreover, many companies that use Delphi, keep quiet about it, don't advertise it in their job ads, etc. Since Delphi-produced apps can compile to a single EXE, run very fast, etc., there is no obvious sign to customers that it's a Delphi app. Many Delphi shops like this, and are happy to let customers and competitors assume they are using something else.
I also used to think, until a few weeks ago, that "2:00am is midnight and 12:00pm is noon". Then someone told me that was wrong. I stubbornly insisted that was right.
You can guess how well that worked out...
I Googled for a while, and read a number of apparently authoritive sources. It appears that "common knowledge" I had was wrong, and the only correct description of those times is noon and midnight. "12 noon" and "12 midnight" are also correct, though the 12 adds no information.
[Smart 'make' routine which can take advantage of compiled object files in other people's views.]
This is brilliant in theory, and I am impressed by it. However, I found it less useful in actual practice. In a Java project with a few thousand classes, and a massive Clearcase server, I found that I could do a full compile of the project, stored outside of CC, in a small fraction of the time of an incremental or full compile in CC, regardless of what other files other users had compiled etc.
* Directory access is really, really slow. Doing an "ls" of an directory in a big project could take many seconds. (With a very expensive and well-equipped server and local network.)
* File access is slow.
* A company with 100 developers and CC, probably has a CC adminsitrator. A company with 100 developers and CVS, probably has a few developers who tend to CVS on a very part-time basis.
* Doesn't have atomic checkin of more than one file, one of the features I also miss in CVS.
* CC gets tangled up with build processes, they have their own "make" and apparently there are issues with using an normal Make in a CC view. I believe such entanglements are unwise, and controlling sources should be isolated from building, such that either could be replaced completely without affecting the other.
* I think the whole model of my "view" on a server, is inferior to having the set of files I need on my local, huge, cheap, very fast hard drive. Trying to beat a $200 hard drive, with a network, pile of complex software, and massive server, is a sucker's bet.
* I prefer a system where the developer says "OK, go get everyone else's changes now" to the dynamic views (sometimes called "Jello views") in ClearCase. (Yes, I know that newer versions of CC have snapshot views.)
* CC has silly installation requirements because of how it works with virtual directories/drives... for example, installing an NT Service on developer *client* machines.
* To reiterate the first points, I found that CC slows down developers. Time is money, and CC costs too much time.
I'm vastly happier with CVS than CC. When some of CVS's issues bother me enough, I will get something else. My minimum requirement for any source control system is that it is not worse than CVS in any important way.
I suppose I should have made that an actual link:
.NET
Slides and notes from a talk on Java/J2EE and
For some reason I naively assumed it would become a link automatically. If only I would learn to preview...
By coincidence, I gave a talk on how Java and .NET compare, just a few days ago:
t ml
.NET has a lot of things "in the box" that a developer / team needs to go find, select, download, install, etc. for Java. I personally like the best-of-breed approach that can result from that (choose from the many SOAP implementation, choose the best one for our needs), but for many developers, having a good-enough solution "in the box" is more palatable.
http://kylecordes.com/story-260-java-net-talk.h
One thing with the SD Times identified as a source of trouble (the size of the included libraries), I identified as a strength:
This is particularly exacerbated in some development shops where extensive format decision and permission processes are needed to make any third party tool purchase (or free software adoption).
Something important to consider is the noise generated by the fan in a projector. There may be some quiet ones out there, but most projectors I encounted, regardless of DLP/LCD/etc., are too loud. (How loud is too loud? I expect a $2-5000 TV to be at least not much worse than a $200 TV...)
The V1 had a single, large chamber (for a smaller number of larger bangs); the ones in this article have a bunch of small chambers, each running many cycles per second, for combined rate of hundreds of smaller bangs for second. Perhaps, once it's scaled up to production size, all these smaller bangs will blend together in to a roar, which might be similar in loudness to a normal get engine of similar size.
Then again, maybe not.
Can you give an example of something that is worth paying for, but not worth paying a full penny for?
I looked at getting a single behemoth display like that, but decided that two smaller screens (two 19" LCDs) are a much better deal - more pixels for less money.
For some uses, having two screens would be a problem (i.e. graphic editing), but for software development it works quite well.
I have two 191T's, and am extremely happy with them. I also leave the brightness turned down pretty far.
Feeding both with DVI, the image quality is far more sharp and steady than any CRT I've used.
The essence of the elections is not whether they are conducted with the same set of rules as in the U.S.; the essence is whether there is some semblance of democracy at all.
:-)
I would not even expect offhand that Iraq will have a governmental structure or election structure like the U.S. has. Keep in mind that the way the U.S. elections, federal/state system work, etc., are a 200 year old legacy system, not something that the current generation of leaders dreamed up
Somewhere along the way, I believe a NASA person pointed out that there is only a brief time just before landing when bailout is possible. The shuttle comes in very fast, still supersonic even partway across Florida.
Not always; there can be a lot of value in an unenforcable patent, to create a chilling effect on competitors, especially the smaller ones. That's because noone really knows if it's enforcable until someone can afford to spend a hefty sum on litigation to find out.
The way to address this is to hire good people, and treat them well, especially for key positions like this one. That way when the person leaves to move on in their career, it will be with regret and good wishes. They will be eager to help you in way they can help you recover from their departure, rather than looking for ways to harm you.
A while back a resigned from a company. I got a big chuckle out of the HR people handing the "termination". The wording of all the documents was as though they were doing the terminating, which was far from the case (the management and I had and still have a good relationship).
There are of course some similarities between firing someone and them leaving to move on in their career, but treating the latter too much like the former, just makes you look silly.
I am very pleased with DreamHost. Recommended.
I have two 191T's in front of me now. They are superb; I haven't noticed any dead pixels yet in either of them, after a couple months of daily use.
I run DVI input on both; for a while I had DVI on one and analog on the other, and the difference was immediately noticable and substantial. Ignore those who claim that DVI in unnecessary; at 1280 resolution, DVI is a large, important, don't-go-without-it difference.
I find 1280 resolution to be quite nice on a desktop 19" LCD. This is probably because I have it positioned at a much farther and more comfortable distance than my 15" laptop display was.
Also 19" LCDs unfortunately cost rather more than $500, for good quality ones anyway.
By using .EXE and .DLL for the various kinds of ready-to-deploy .NET code, Microsoft helped enormously in making it feel native to those who develop and deploy it. You could deploy an app, consisting of an EXE and some DLLs, without even noticing that it's .NET based, particular down the road a bit when the .NET runtime is more widely deployed.
.NET feel native, in a way that Java apps do not. It's not a technical difference, it's a cosmetic difference that will make adoption a bit easier.
This naming is one of several things that will make
I'm working under the assumption that their errant software update did not physically damage the cable modem, it just put in some updated software that renders it unusable.
I suggest contacting the manufacturer of the cable modem; there may be a combination of some "reset" step and some kind of firmware update that would overwrite whatever badness the cable company (accidentally, I assume) put on there, and get it going again.
I've recently switched from SpamAssasin to POPFile for the same reason. With SA, once the version I had became slighly outdated, spammers had found ways to write spam that didn't set it off, and a steady flow got through. With POPFile, I generally only ever see one or two of a new flavor of spam, since after a couple of reclassifications the rest gets detected.
IP law is expensive and time consuming. I doubt you will get any assurance of anything from an IP lawyer for $600.
This is not much of an insight, but:
I'd say the probability is about 95% that this idea is going nowhere, and 5% that it's worth $10.
Delphi is not very popular for "enterprise" development. However, it is quite popular for shrinkwrap software development, especially vertical market apps. Moreover, many companies that use Delphi, keep quiet about it, don't advertise it in their job ads, etc. Since Delphi-produced apps can compile to a single EXE, run very fast, etc., there is no obvious sign to customers that it's a Delphi app. Many Delphi shops like this, and are happy to let customers and competitors assume they are using something else.
It seems to me that it would take a rather foolish web designer / programmer / etc. to make sure a guarantee.
I also used to think, until a few weeks ago, that "2:00am is midnight and 12:00pm is noon". Then someone told me that was wrong. I stubbornly insisted that was right.
You can guess how well that worked out...
I Googled for a while, and read a number of apparently authoritive sources. It appears that "common knowledge" I had was wrong, and the only correct description of those times is noon and midnight. "12 noon" and "12 midnight" are also correct, though the 12 adds no information.
[Smart 'make' routine which can take advantage of compiled object files in other people's views.]
This is brilliant in theory, and I am impressed by it. However, I found it less useful in actual practice. In a Java project with a few thousand classes, and a massive Clearcase server, I found that I could do a full compile of the project, stored outside of CC, in a small fraction of the time of an incremental or full compile in CC, regardless of what other files other users had compiled etc.
My thoughts on ClearCase:
* Directory access is really, really slow. Doing an "ls" of an directory in a big project could take many seconds. (With a very expensive and well-equipped server and local network.)
* File access is slow.
* A company with 100 developers and CC, probably has a CC adminsitrator. A company with 100 developers and CVS, probably has a few developers who tend to CVS on a very part-time basis.
* Doesn't have atomic checkin of more than one file, one of the features I also miss in CVS.
* CC gets tangled up with build processes, they have their own "make" and apparently there are issues with using an normal Make in a CC view. I believe such entanglements are unwise, and controlling sources should be isolated from building, such that either could be replaced completely without affecting the other.
* I think the whole model of my "view" on a server, is inferior to having the set of files I need on my local, huge, cheap, very fast hard drive. Trying to beat a $200 hard drive, with a network, pile of complex software, and massive server, is a sucker's bet.
* I prefer a system where the developer says "OK, go get everyone else's changes now" to the dynamic views (sometimes called "Jello views") in ClearCase. (Yes, I know that newer versions of CC have snapshot views.)
* CC has silly installation requirements because of how it works with virtual directories/drives... for example, installing an NT Service on developer *client* machines.
* To reiterate the first points, I found that CC slows down developers. Time is money, and CC costs too much time.
I'm vastly happier with CVS than CC. When some of CVS's issues bother me enough, I will get something else. My minimum requirement for any source control system is that it is not worse than CVS in any important way.