I work for a small, independent software vendor and one of my more tedious tasks is to package new releases of our product for the platforms we support. I've grown to hate doing it. I create pkgs for Solaris (SPARC and Intel), Linux (RedHat RPM only, haven't got time for anything else) and others. (Our NT guys build InstallShield packages.) Some, like AIX and FreeBSD, we still haven't figured out and have resorted to tar files. I have to build the packages natively on each platform, usually from binaries installed in-place and common files from RCS, and then drag the results over to a central media directory from which we cut our CDs (or upload them to our web site). The process is error-prone, long-winded and non-trivial, even with Makefile help, etc. Then I have to document different installation, upgrade and removal processes for each platform.
Obviously, when the prospect of supporting a new platform is raised, I groan inwardly. No actually, I groan loudly too. As for supporting the many different variations of platforms (e.g. Alpha, SPARC, PowerPC, Debian/Caldera, etc.) - FORGET IT! It would take a month just to get a new release out of the door. (Obviously, there are also issues of available hardware, development, support staff, etc.)
I see a lot of commercial Unix software that is supplied in tar or cpio format with a customised install script, so obviously others have opted to circumvent the above hassle entirely. However, this way you lose the benefits of the system's own software management, versioning, etc.
I like RPM myself as a user, but it doesn't solve my problems as a packager because it's not standard. And heck, why should anyone bother fixing or improving its deficiencies when they can just write their own packager?! Yeah, make that wheel rounder! Sorry, I'm starting to sound bitter...
In an ideal world, software would come in source format and there would be an easy config/compile/install tool for all of it. In practice, it ain't never gonna happen. Even if we did bravely decide to release open source our products, at present, I wouldn't trust all our users to have a working compiler, a correct set of libraries or necessarily, a clue.
A good package manager supplies:
easy install/upgrade (while preserving local user files)/uninstall, either interactive or automated;
a central database of installed software, to query installed versions, contents, etc.;
dependency management;
conflict management ("this file overwrites file from package X");
automated package building for developers, with copious documentation;
open package format and API;
A really good package manager would also work on any platform. I don't think one can count on vendors like Sun to support a new format unless it becomes either a formal, popular or defacto standard (hey, they have eventually bundled gzip and patch!). Therefore, executable installers (consisting of the software plus a standard wrapper for the platform in question) would seem to be required. In turn, this infers the need for an easy way for developers to build such packages on any single platform, requiring only the pre-built platform-specific parts (binaries and libraries) from elsewhere.
Ghod, I'd love that.
(Alternatively, tools that convert, create or unpack multiple foreign package formats would be nice. I'll have a look at PkgMaker.)
I concur. By claiming that this was the most important and memorable moment of a lifetime, the guy is buying into the same media hype he decries.
None of the other events listed fall into the same category. Things like that can occur any time: we may yet put a man on Mars, for example. Hostage crisis? Uh, see the news from Afghanistan last week? And anyone who seriously needs the thrill of a presidential assassination can always arrange it themselves (as for resignation, didn't Clinton come close enough for ya?).
Yes, it was hyped. Yes, I worked. WTF, there was nothing else worth doing and I wanted the cash. OK, it sucks if you're ordered to work when you'd rather be at a party, but it's not like there aren't other parties and other jobs.
Cronolog can be used on the end of a pipe from Apache (or presumably anything that generates logs similarly) and will automatically write logs to paths keyed on date. E.g. if you want to collect each month's logs in separate dirs, cronolog will write to 1999/Oct, 1999/Nov, etc. It's an extremely useful way of splitting up your log files chronologically without writing scripts to restart Apache and move the old logs. Do a search on Google or somewhere for it.
Au contraire, I have been down this singularly unrewarding path before and have assembled all the info I found on DDS/DAT here. That includes advice, links, READMEs, software, threads, searches, firmware, etc. In summary: you can do it, but it ain't easy. Finding a suitable DDS drive is the biggest challenge. If you have any more, let me know.
The privacy violations are obviously of some concern, but the only significant difference from countless other such marketing ploys that I can see is that Amazon is publishing this data publically. And ironically, it's data on the very same companies that would seek to follow the same practice! Every company wishes to build up these kind of buyer profiles so that they can be even nosier and bother us more often (targetted bothering only, of course;-).
Heck, at least Amazon are coming (partially) clean about what data they hold and how they use it. And they're doing it with the big names that do it to us in secret. Excuse me while I weep crocodile tears.
This is not to say I want to go down the slippery path towards full disclosure of individual purchases, or even that Amazon's "cool feature" is a good thing long term. But the action itself is quite radical in this form.
So the difference between the Linux crowd and the mainstream is that they use half-naked chicks to promote their wares at trade shows and we use innocent animals? I'm sooooo proud...
This is an idea whose time should have come a few years ago and indeed which has been anticipated by the journal publishers since they became aware of the Internet. I worked for one of the largest scientific publishers early in my career. From the sidelines in IT support, my impression was that the business "strategists" (I use the word loosely) were deeply worried that the Internet would allow the scientific community to bypass them and were desperately scrambling to find a way of harnessing the net that would still keep them in the loop. This took the form of multiple electronic publishing projects with few overarching goals or standards, offering little added value to customers and heading nowhere.
The community has the tools to take on the role of the publishers themselves - they're the same ones the publishers are using (LaTeX, SGML, Perl, Apache...). What they lack so far is the organisational infrastructure to recreate the peer review and editorial process (I assume that these are still required). It is interesting to finally hear rumbles of discontent, motivation and progress.
Should they succeed in the DIY approach, it sounds as if the publishers will have no one to blame but themselves - a situation analogous to the record companies vs. the online world. Scientific publishing has been a licence to print money for years: Robert Maxwell funded the expansion of his empire through revenue from subscriptions to Pergamon Press journals (one of the few industries where customers pay a year in advance of product delivery, which occurs in installments!).
This is a little off-topic as the original question was about GUI designer packages (not whether whether vi was better, chaps!). However, if you want to generate large amounts of HTML pages quickly from standard templates and have the facility to regenerate it after tweaking the template, other posters are right in recommending automation. (Personally, I hate any package that requires me to move my hands from the mouse to the keyboard more than once per minute.:-)
I've used the m4 macro techniques outlined in the following references and find them excellent for standardising pages and removing the worst pains of handcoded HTML. Every day, I find new ways to extend them. The downsides are coping with m4's syntax requirements (mind your quotes!) and the initial work creating your macros. If this doesn't suit, try some of the many other HTML preprocessing utils (see Freshmeat).
As an aside, some of the nicest pages I've seen used extremely effective graphics way beyond what I could draw - but were a pain to load and use. I've seen simpler sites that did nifty things with TABLE layouts instead.
Stern's comments about the use of IT actually retarding productivity in the last thirty years sound like an uncanny echo of Thomas K. Landauer's thesis in "The Trouble With Computers: Usefulness, Usability, and Productivity". Haven't yet managed to finish this book (otherwise a review would be most appropriate, I guess;-), but it hammers the point home and even offers a few solutions. NB. It's nothing like Dave Barry.
When you're on course to take the fall for managerial incompetence/stupidity/arrogance. (Ie. when they're not listening.)
When your quality of life outside work is poor (lack of social life, bad environment, etc.)
When the organisation doesn't learn from its mistakes.
When your future prospects are limited, either by design or simply the size of the organisation.
When you're BORED.
NB. It helps to realise what's good about your present employment, and then you can either look for those factors next time or at least accept change with a clear idea of what you'll be losing. I found a list of pros and cons, with scores, immensely helpful in coming to a decision. Finally, always bear in mind that you can leave at any time.
"...We preceive and yet interweave, casting ever-contemporaneous shadows of metaphor upon the bleak technosphere of the digital colossus. Slowly, our mental experience folds in upon us, and yet expands, filling our consciousness with half-remembered idioms of symbolic platitudes. We have become, and yet remain who we are, but then unlike. The knowledge sears me, manipulating chaotic pools of raw being to produce understanding within only ignorance, and this nexus connects interminably towards a higher belief of self self SELF."
(Anyone can write bollocks. Perl is more useful.) For chrissakes, how much longer do we have to entertain these old school charlatans?? If I have to read one more clueless protozoan hack struggling to get an angle on what is, in the end, just a stream of bits, albeit a bloody useful one, I will hurl. And actually, I won't have to read it on Slashdot because Richard's joining Katz in the killfile, but it offends me that anyone offers these pretentious bozos pagespace. I for one am not consensual to sharing my space, perceived or otherwise, with this vacuous prattling.
Strikes me that one could advance the case for Linux as an enterprise platform much further if Veritas could be persuaded to port Vxfs and Volume Manager to it. Lack of journalled filesystem support (beyond some experimental stuff in the very early stages) is a major drawback to Linux as a highly available, robust platform. And Vxfs would supply a common, portable filesystem format.
Customers like Oracle on Solaris because Sun can point to hundreds of reference sites, many of which were installed by themselves in short periods. Buy the hardware (servers, arrays); buy the software (OS, disk management, backup); plug it all in; configure; it works. (Yes, you need to have done it before.) In contrast, Linux often seems to require a degree of tinkering by a knowledgeable admin to get the best from it.
The fact that ESR didn't look a little harder at the URL and work out that Thomas Scoville wrote the article suggests his response was written quickly in a less understanding mood than he might otherwise be in. (Yes, Scoville should have included an attribution on the original page, but I guess this was an oversight.)
Of course business can't "crush" OSS in any real sense, short of social domination. It may be able to marginalise it, but so long as we can download the software we want from the net, I doubt that will have much impact either. But it's still fair to say that business doesn't "get" OSS. Offering bounties and investing in the companies perceived to be the leading players are illustrations of business "joining" the OSS movement the only way it knows how: by spending money for something tangible.
Paying a programmer to write software only the business wants and then releasing the source will not realise the true benefits of OSS. Those occur when the programmer writes software he or she wants. (Pop quiz: name one leading female OSS programmer or explain why you can't. Worrying?) Where business can help is in paying those programmers so they don't otherwise lose time earning a living by undertaking unrelated activities. Unfortunately, that's only likely to happen where the programmer coincidentally happens to be creating something that the business wants. Red Hat *needs* a good desktop; it so happens that this is sexy stuff to many coders anyway, and they found people already working on a suitable package. IBM may have to look harder to find the folks coding enterprise features, but I'm sure they exist. The danger is that those projects succeed at the expense of related projects following other designs; this introduces imbalances into OSS output and potentially leads to software monopolies, although probably of better software than that provided by the proprietory world (eg. GNOME becoming the best-supported desktop).
I use Mutt at work and Eudora on my Mac at home. You're comparing apples to oranges. I love mutt but...if you just want to read mail rather than configure how to read it, Eudora is very, very, very nice. Linux could use a GUI client that's so elegant and simple, particularly if it is to pay more than lip-service to the idea of user-friendliness.
I've used a NatWest cash till here in the UK with a similar NT error dialogue on it - something about not being able to fetch a current profile. It still worked OK (unless Bill now has a direct debit on my account:-) but the box obscured the display. Of course, NatWest has had well-publicised "issues" with NT. Serve 'em right. Errors like this on a customer-facing application are unforgiveable.
- Ignore the stuff about eat/sleep/breathe Unix. If you like this stuff enough to be a sysadmin, you'll do it in your spare time anyway. (After a few years in a related job, you probably won't want to anymore though.) I suggest you remind yourself what summer is occasionally.
- CS degree of whatever sort is useful, if only to let you play on a decent network. In the UK, I think academic CS depts have acted as a useful training ground for many sysadmins. Whether this is still true in the age of public spending cutbacks, I don't know, but I certainly did my apprenticeship in that environment by staying on in Support after my degree. For the first couple of years after graduation, I considered my CS syllabus to be completely irrelevant to my work. I now see that an overview of subjects such as project management and documentation (my degree had a heavy, large-scale software engineering emphasis) was extremely useful and I wish I'd paid even more attention. Many sysadmins are now expected to act as the implementors for new business systems in companies. This kind of work shades into consultancy. You can do it without the degree, but it's different again. Those who have will always claim you don't need one, and vice-versa. It would be nice if universities cottoned on and started offering sysadmin modules as options in their course syllabi.
- Yes, being able to communicate in TCP is important, but people skills are equally so. Too many admins are viewed as rude, arrogant, pointless, unhelpful and/or egotistical by their users (and I'm not trying to evade guilt myself).
- Join USENIX & SAGE if you care about this profession.
Good luck with your intended career, it's an excellent choice.
Slashdot Mirror
Obviously, when the prospect of supporting a new platform is raised, I groan inwardly. No actually, I groan loudly too. As for supporting the many different variations of platforms (e.g. Alpha, SPARC, PowerPC, Debian/Caldera, etc.) - FORGET IT! It would take a month just to get a new release out of the door. (Obviously, there are also issues of available hardware, development, support staff, etc.)
I see a lot of commercial Unix software that is supplied in tar or cpio format with a customised install script, so obviously others have opted to circumvent the above hassle entirely. However, this way you lose the benefits of the system's own software management, versioning, etc.
I like RPM myself as a user, but it doesn't solve my problems as a packager because it's not standard. And heck, why should anyone bother fixing or improving its deficiencies when they can just write their own packager?! Yeah, make that wheel rounder! Sorry, I'm starting to sound bitter...
In an ideal world, software would come in source format and there would be an easy config/compile/install tool for all of it. In practice, it ain't never gonna happen. Even if we did bravely decide to release open source our products, at present, I wouldn't trust all our users to have a working compiler, a correct set of libraries or necessarily, a clue.
A good package manager supplies:
A really good package manager would also work on any platform. I don't think one can count on vendors like Sun to support a new format unless it becomes either a formal, popular or defacto standard (hey, they have eventually bundled gzip and patch!). Therefore, executable installers (consisting of the software plus a standard wrapper for the platform in question) would seem to be required. In turn, this infers the need for an easy way for developers to build such packages on any single platform, requiring only the pre-built platform-specific parts (binaries and libraries) from elsewhere.
Ghod, I'd love that.
(Alternatively, tools that convert, create or unpack multiple foreign package formats would be nice. I'll have a look at PkgMaker.)
Ade_
/
I concur. By claiming that this was the most important and memorable moment of a lifetime, the guy is buying into the same media hype he decries.
None of the other events listed fall into the same category. Things like that can occur any time: we may yet put a man on Mars, for example. Hostage crisis? Uh, see the news from Afghanistan last week? And anyone who seriously needs the thrill of a presidential assassination can always arrange it themselves (as for resignation, didn't Clinton come close enough for ya?).
Yes, it was hyped. Yes, I worked. WTF, there was nothing else worth doing and I wanted the cash. OK, it sucks if you're ordered to work when you'd rather be at a party, but it's not like there aren't other parties and other jobs.
Ade_
/
Cronolog can be used on the end of a pipe from Apache (or presumably anything that generates logs similarly) and will automatically write logs to paths keyed on date. E.g. if you want to collect each month's logs in separate dirs, cronolog will write to 1999/Oct, 1999/Nov, etc. It's an extremely useful way of splitting up your log files chronologically without writing scripts to restart Apache and move the old logs.
Do a search on Google or somewhere for it.
Ade_
/
Au contraire, I have been down this singularly unrewarding path before and have assembled all the info I found on DDS/DAT here. That includes advice, links, READMEs, software, threads, searches, firmware, etc.
In summary: you can do it, but it ain't easy. Finding a suitable DDS drive is the biggest challenge.
If you have any more, let me know.
Ade_
/
Slashdot as a community? Anecdotes from your time "in" Slashdot? Friends you have made??
Send for Katz! He's their ideal subject. They deserve one another!
Ade_
/
(Does he still write for Slashdot? I wouldn't know, I filter his stuff.)
The privacy violations are obviously of some concern, but the only significant difference from countless other such marketing ploys that I can see is that Amazon is publishing this data publically. And ironically, it's data on the very same companies that would seek to follow the same practice! Every company wishes to build up these kind of buyer profiles so that they can be even nosier and bother us more often (targetted bothering only, of course ;-).
Heck, at least Amazon are coming (partially) clean about what data they hold and how they use it. And they're doing it with the big names that do it to us in secret. Excuse me while I weep crocodile tears.
This is not to say I want to go down the slippery path towards full disclosure of individual purchases, or even that Amazon's "cool feature" is a good thing long term. But the action itself is quite radical in this form.
Ade_
/
So the difference between the Linux crowd and the mainstream is that they use half-naked chicks to promote their wares at trade shows and we use innocent animals? I'm sooooo proud...
Ade_
/
This is an idea whose time should have come a few years ago and indeed which has been anticipated by the journal publishers since they became aware of the Internet. I worked for one of the largest scientific publishers early in my career. From the sidelines in IT support, my impression was that the business "strategists" (I use the word loosely) were deeply worried that the Internet would allow the scientific community to bypass them and were desperately scrambling to find a way of harnessing the net that would still keep them in the loop. This took the form of multiple electronic publishing projects with few overarching goals or standards, offering little added value to customers and heading nowhere.
The community has the tools to take on the role of the publishers themselves - they're the same ones the publishers are using (LaTeX, SGML, Perl, Apache...). What they lack so far is the organisational infrastructure to recreate the peer review and editorial process (I assume that these are still required). It is interesting to finally hear rumbles of discontent, motivation and progress.
Should they succeed in the DIY approach, it sounds as if the publishers will have no one to blame but themselves - a situation analogous to the record companies vs. the online world. Scientific publishing has been a licence to print money for years: Robert Maxwell funded the expansion of his empire through revenue from subscriptions to Pergamon Press journals (one of the few industries where customers pay a year in advance of product delivery, which occurs in installments!).
Ade_
/
Ade_
/
This is a little off-topic as the original question was about GUI designer packages (not whether whether vi was better, chaps!). However, if you want to generate large amounts of HTML pages quickly from standard templates and have the facility to regenerate it after tweaking the template, other posters are right in recommending automation. (Personally, I hate any package that requires me to move my hands from the mouse to the keyboard more than once per minute. :-)
I've used the m4 macro techniques outlined in the following references and find them excellent for standardising pages and removing the worst pains of handcoded HTML. Every day, I find new ways to extend them. The downsides are coping with m4's syntax requirements (mind your quotes!) and the initial work creating your macros. If this doesn't suit, try some of the many other HTML preprocessing utils (see Freshmeat).
As an aside, some of the nicest pages I've seen used extremely effective graphics way beyond what I could draw - but were a pain to load and use. I've seen simpler sites that did nifty things with TABLE layouts instead.
Ade_
/
Does anyone want to buy my Microdrive?
Ade_
/
Stern's comments about the use of IT actually retarding productivity in the last thirty years sound like an uncanny echo of Thomas K. Landauer's thesis in "The Trouble With Computers: Usefulness, Usability, and Productivity". Haven't yet managed to finish this book (otherwise a review would be most appropriate, I guess ;-), but it hammers the point home and even offers a few solutions. NB. It's nothing like Dave Barry.
NB. It helps to realise what's good about your present employment, and then you can either look for those factors next time or at least accept change with a clear idea of what you'll be losing.
I found a list of pros and cons, with scores, immensely helpful in coming to a decision.
Finally, always bear in mind that you can leave at any time.
Ade_
/
(Anyone can write bollocks. Perl is more useful.)
For chrissakes, how much longer do we have to entertain these old school charlatans?? If I have to read one more clueless protozoan hack struggling to get an angle on what is, in the end, just a stream of bits, albeit a bloody useful one, I will hurl.
And actually, I won't have to read it on Slashdot because Richard's joining Katz in the killfile, but it offends me that anyone offers these pretentious bozos pagespace. I for one am not consensual to sharing my space, perceived or otherwise, with this vacuous prattling.
Strikes me that one could advance the case for Linux as an enterprise platform much further if Veritas could be persuaded to port Vxfs and Volume Manager to it. Lack of journalled filesystem support (beyond some experimental stuff in the very early stages) is a major drawback to Linux as a highly available, robust platform. And Vxfs would supply a common, portable filesystem format.
Customers like Oracle on Solaris because Sun can point to hundreds of reference sites, many of which were installed by themselves in short periods. Buy the hardware (servers, arrays); buy the software (OS, disk management, backup); plug it all in; configure; it works. (Yes, you need to have done it before.) In contrast, Linux often seems to require a degree of tinkering by a knowledgeable admin to get the best from it.
Ade_
/
The fact that ESR didn't look a little harder at the URL and work out that Thomas Scoville wrote the article suggests his response was written quickly in a less understanding mood than he might otherwise be in. (Yes, Scoville should have included an attribution on the original page, but I guess this was an oversight.)
Of course business can't "crush" OSS in any real sense, short of social domination. It may be able to marginalise it, but so long as we can download the software we want from the net, I doubt that will have much impact either. But it's still fair to say that business doesn't "get" OSS. Offering bounties and investing in the companies perceived to be the leading players are illustrations of business "joining" the OSS movement the only way it knows how: by spending money for something tangible.
Paying a programmer to write software only the business wants and then releasing the source will not realise the true benefits of OSS. Those occur when the programmer writes software he or she wants. (Pop quiz: name one leading female OSS programmer or explain why you can't. Worrying?) Where business can help is in paying those programmers so they don't otherwise lose time earning a living by undertaking unrelated activities. Unfortunately, that's only likely to happen where the programmer coincidentally happens to be creating something that the business wants. Red Hat *needs* a good desktop; it so happens that this is sexy stuff to many coders anyway, and they found people already working on a suitable package. IBM may have to look harder to find the folks coding enterprise features, but I'm sure they exist. The danger is that those projects succeed at the expense of related projects following other designs; this introduces imbalances into OSS output and potentially leads to software monopolies, although probably of better software than that provided by the proprietory world (eg. GNOME becoming the best-supported desktop).
Ade_
/
...That well-known colloquialism. ;-/
I use Mutt at work and Eudora on my Mac at home. You're comparing apples to oranges. I love mutt but...if you just want to read mail rather than configure how to read it, Eudora is very, very, very nice. Linux could use a GUI client that's so elegant and simple, particularly if it is to pay more than lip-service to the idea of user-friendliness.
Ade_
/
I've used a NatWest cash till here in the UK with a similar NT error dialogue on it - something about not being able to fetch a current profile. It still worked OK (unless Bill now has a direct debit on my account :-) but the box obscured the display. Of course, NatWest has had well-publicised "issues" with NT. Serve 'em right. Errors like this on a customer-facing application are unforgiveable.
A few more remarks:
- Ignore the stuff about eat/sleep/breathe Unix. If you like this stuff enough to be a sysadmin, you'll do it in your spare time anyway. (After a few years in a related job, you probably won't want to anymore though.) I suggest you remind yourself what summer is occasionally.
- CS degree of whatever sort is useful, if only to let you play on a decent network. In the UK, I think academic CS depts have acted as a useful training ground for many sysadmins. Whether this is still true in the age of public spending cutbacks, I don't know, but I certainly did my apprenticeship in that environment by staying on in Support after my degree.
For the first couple of years after graduation, I considered my CS syllabus to be completely irrelevant to my work. I now see that an overview of subjects such as project management and documentation (my degree had a heavy, large-scale software engineering emphasis) was extremely useful and I wish I'd paid even more attention. Many sysadmins are now expected to act as the implementors for new business systems in companies. This kind of work shades into consultancy.
You can do it without the degree, but it's different again. Those who have will always claim you don't need one, and vice-versa. It would be nice if universities cottoned on and started offering sysadmin modules as options in their course syllabi.
- Yes, being able to communicate in TCP is important, but people skills are equally so. Too many admins are viewed as rude, arrogant, pointless, unhelpful and/or egotistical by their users (and I'm not trying to evade guilt myself).
- Join USENIX & SAGE if you care about this profession.
Good luck with your intended career, it's an excellent choice.