In a summary you can state your conclusions as assertions, without the explanation and justification you include later on. I would have thought that writing some bold statements in the "executive summary" would help to capture the reader's attention and interest and so increase the chances of his/her reading the whole document.
He didn't invent the name, but it was surely his choice whether to use it or not. The details aren't important.
2. The problem of moving applications from 32 to 64 bits is NOT the instruction set, but the high-level language used. Many nontrivial C and C++ programs cannot be just compiled in 64 bits w/o some modifications (and they are not so dirty).
It is the instruction set, indirectly. The "long long" type is a GNU extension to Standard C (though I think it is in the C9X draft standard), so time_t must be defined as just plain long, which is 32-bit.
3. Does not address the issue of packaging incompatibility among distributions,
Not specifically. It covers the differences between distributions very briefly.
and lack of means of verifying that an installation has not been modified from an approved setup (otherwise, it can become a tech support nightmare).
Are rpm --verify and dpkg --audit not good enough for you?
There are a lot of errors in this, and I think the tone is quite wrong for giving to a manager. However, I think that constructive criticism to the author and some help with editing could result in a much better document in a few weeks.
That paper is written by some of my colleagues. It's possible that I could ask them to help, if you explain what they can do. I've emailed you more info, assuming that you receive email to big_head_fred@hotmail.com.
Maybe you didn't read the lot. How's about this FUD:
Time for a reality check, folks! No broad market acceptance. Stigmatized as freeware. No support. No training or certification. No network management tools. Unproven in large corporations. Other than that, Linux is a fine product.
I'm not sure what is meant by broad market acceptance, but I certainly don't think Linux is just filling a niche. The "stigma" of freeware has all but disappeared. Contracted support is there; training is available; as is certification for administration on specific distributions. Some network management tools are there - I'm not sure what in particular they might be interested in. Several large corporations are successfully using Linux (not for "mission-critical" stuff, but I don't see that qualification given). This is just the mixture of untruths, exaggerations, and half-truths that we've come to expect from people that don't want to accept Linux.
The fact that Linux can't yet provide really high availability remains, but that doesn't justify this.
The legal system is like an operating system. Court procedures define the scheduler and resource allocation. Case-law is available to you as subroutines. A good legal argument is a good hack. Erm, I think. I did talk this through with a law student a while back.
The high school I used to goto in Toronto, Canada has full, high speed Internet. This thing started where everytime a lab was left unattended, people would pull up a porno site on all the computers in the lab. When the teachers, or visitors came into the lab... you get the picture. They tried to setup a proxy, but they couldn't figure out how to make it work.
Why pay all that money for high-speed access and then not get a proxy? Caching proxies not only allow logging and blocking, but can speed up access. I don't think the blocking is desirable or necessary. Logging, however, means that if anyone pulls a stunt like they can be tracked down and disciplined. If students know that, they won't do it.
The envelope is the information used in SMTP. As far as SMTP is concerned, the headers and body are one long stream of bytes. The mail server has to decide whether or not to accept a message before it sees the headers.
Sendmail, bind, and the BSD networking code have been around for a *very* long time, and there have been no "incompatible" splintering.
The article didn't talk about incompatible splintering but simply forking. BSD has forked many times, and the BSD networking code even more. The BSD license sort of encourages this by allowing the creation of proprietary versions whose source cannot be folded back into a mainstream version.
He didn't say he exepcted to be able to get anything out in 6 months; what he said is that ideally Netscape should have done so - in order to maintain it's mindshare, I suppose, if nothing else. If you read the letter thoroughly you will see that he acknowledged that this was never practical.
No, it says "raped or harassed" (my emphasis. I suspect that rape accounts for a quite small fraction of that 1/6. That makes it quite a misleading statistic. How about separating the two?
Ticket scalpers? People who chop off the top of tickets' heads? Oh, you mean ticket touts.
Linux needs formal engaged testing? Don't be sure.
on
Slate Takes on Linux
·
· Score: 1
At the moment, Linux has no formal regression testing, and there's a lot of design stuff that isn't documented. There is certainly the potential for serious breakage, and sometimes it does happen. Regression testing of "stable" preX versions might be very helpful in eliminating the bugs that occasionally pop up in the stable series. That said, Linux's design is mostly quite robust and Linus keeps very tight quality control over the core.
The parallelised nature of "bazaar" development probably reduces the cost of fixing bugs, because it's easier to create and test branches off the official development versions.
The "pale and wan" reference was surely a reference to fvwm95, in which case it would be completely justified. I don't know why Red Hat makes that crap the default.
Linus doesn't think freedom is important - like ESR, he thinks practical gains are more important.
I'm sure Linus considers the freedoms of free software to be useful and important to its success. I think the difference is that he and ESR don't consider there to be a moral issue here.
Seriously, people who know something are often percieved by thos who lack the same knowledge as arrongant.
That's because they often are arrogant.
Slashdotters as a whole are very knowledgable people. Linux users in particular are either very knowledgable, or at least fully engaged in the often painful process of becoming so.
Surely you jest! Certainly there are many Slashdot users that are very knowledgeable about computer technology. I think they are a minority now. There are a lot of intemperate, ignorant, illiterate users that have little to say but make sure to say it anyway. (e.g. the absurd calls for software to be made free right now, or the equally absurd calls for people to shut up about licenses.)
We have this reinforced nearly every day, with our phones ringing off the hook with people desperate to hire us away from our current employers.
I think not, especially if you include the nearly 50% of users that are still students.
Arrogant? Maybe. But at least we have something legitimate to be arrogant about. (Not that arrogance is an attractive attribute to have, no matter how well justified).
So you admit it here! It is easy to become arrogant when one does know a lot more about something than most other people. It's never really justified.
A wise, helpful, and patient guru is worth 10,000 of us who are irritable, impatient, and insulting
Quite true. That's why it's best to have a local friend help you to get started with Linux or other "alternative" OS.
Slashdot Mirror
WebDAV is not a specific program. WebDAV is the name given to a particular set of extensions to HTTP that allow for standardised remote management.
In a summary you can state your conclusions as assertions, without the explanation and justification you include later on. I would have thought that writing some bold statements in the "executive summary" would help to capture the reader's attention and interest and so increase the chances of his/her reading the whole document.
He didn't invent the name, but it was surely his choice whether to use it or not. The details aren't important.
It is the instruction set, indirectly. The "long long" type is a GNU extension to Standard C (though I think it is in the C9X draft standard), so time_t must be defined as just plain long, which is 32-bit.
Not specifically. It covers the differences between distributions very briefly.
Are rpm --verify and dpkg --audit not good enough for you?
There are a lot of errors in this, and I think the tone is quite wrong for giving to a manager. However, I think that constructive criticism to the author and some help with editing could result in a much better document in a few weeks.
That paper is written by some of my colleagues. It's possible that I could ask them to help, if you explain what they can do. I've emailed you more info, assuming that you receive email to big_head_fred@hotmail.com.
Maybe you didn't read the lot. How's about this FUD:
I'm not sure what is meant by broad market acceptance, but I certainly don't think Linux is just filling a niche. The "stigma" of freeware has all but disappeared. Contracted support is there; training is available; as is certification for administration on specific distributions. Some network management tools are there - I'm not sure what in particular they might be interested in. Several large corporations are successfully using Linux (not for "mission-critical" stuff, but I don't see that qualification given). This is just the mixture of untruths, exaggerations, and half-truths that we've come to expect from people that don't want to accept Linux.
The fact that Linux can't yet provide really high availability remains, but that doesn't justify this.
Doesn't the Hippocratic Oath also mention not carrying out abortions?
The legal system is like an operating system. Court procedures define the scheduler and resource allocation. Case-law is available to you as subroutines. A good legal argument is a good hack. Erm, I think. I did talk this through with a law student a while back.
Why pay all that money for high-speed access and then not get a proxy? Caching proxies not only allow logging and blocking, but can speed up access. I don't think the blocking is desirable or necessary. Logging, however, means that if anyone pulls a stunt like they can be tracked down and disciplined. If students know that, they won't do it.
The envelope is the information used in SMTP. As far as SMTP is concerned, the headers and body are one long stream of bytes. The mail server has to decide whether or not to accept a message before it sees the headers.
The Mozilla build process has been changed to use autoconf and automake already.
The article didn't talk about incompatible splintering but simply forking. BSD has forked many times, and the BSD networking code even more. The BSD license sort of encourages this by allowing the creation of proprietary versions whose source cannot be folded back into a mainstream version.
Does Consumer Reports ever cover the dangers of carrying out tests for Consumer Reports?
I'm told he had a grilfiend once.
He didn't say he exepcted to be able to get anything out in 6 months; what he said is that ideally Netscape should have done so - in order to maintain it's mindshare, I suppose, if nothing else. If you read the letter thoroughly you will see that he acknowledged that this was never practical.
Shurely 1149 is the best.
No, it says "raped or harassed" (my emphasis. I suspect that rape accounts for a quite small fraction of that 1/6. That makes it quite a misleading statistic. How about separating the two?
Do a Usenet search for articles dated in August 1995 containg the text "Hamilton 95". Extrapolate by 5 years.
Ticket scalpers? People who chop off the top of tickets' heads? Oh, you mean ticket touts.
At the moment, Linux has no formal regression testing, and there's a lot of design stuff that isn't documented. There is certainly the potential for serious breakage, and sometimes it does happen. Regression testing of "stable" preX versions might be very helpful in eliminating the bugs that occasionally pop up in the stable series. That said, Linux's design is mostly quite robust and Linus keeps very tight quality control over the core.
The parallelised nature of "bazaar" development probably reduces the cost of fixing bugs, because it's easier to create and test branches off the official development versions.
The "pale and wan" reference was surely a reference to fvwm95, in which case it would be completely justified. I don't know why Red Hat makes that crap the default.
That should be "GNU/SunOS" since the kernel of Solaris is SunOS.
I'm sure Linus considers the freedoms of free software to be useful and important to its success. I think the difference is that he and ESR don't consider there to be a moral issue here.
That's because they often are arrogant.
Surely you jest! Certainly there are many Slashdot users that are very knowledgeable about computer technology. I think they are a minority now. There are a lot of intemperate, ignorant, illiterate users that have little to say but make sure to say it anyway. (e.g. the absurd calls for software to be made free right now, or the equally absurd calls for people to shut up about licenses.)
I think not, especially if you include the nearly 50% of users that are still students.
So you admit it here! It is easy to become arrogant when one does know a lot more about something than most other people. It's never really justified.
Quite true. That's why it's best to have a local friend help you to get started with Linux or other "alternative" OS.
"Open Source" is a stupid name anyway.
I don't care to use Apple's software. I do care that they are lying about what its license really says. Licenses do matter.
Anti-aliased text? Window management that doesn't break down when your WM gets swapped out?