A follow-up here for the record. I put in a request, and it was granted almost immediately; the DUL guys poked around a bit and when it checked out they went ahead and took my IP out of the DUL jail. I take back what I said. Well, most of it anyway, except for the part about the tone of their Web page, which I still think sounds pretty hostile. Nice to know, though, that they're more reasonable than they sound. --Bob
Well, that's encouraging. I did not do so because after reading through all of the information at http://maps.vix.com/dul/, I could find no indication that such a request would be honored. On the page of information for end users, http://maps.vix.com/dul/enduser.htm, the exclusive remedy is to use your ISP's relay. On the page on removing your network, http://maps.vix.com/dul/removing.htm, the acceptable reasons listed speak only to the needs of ISPs, not individuals. The closest thing I could find was the clause on "Removal due to operational requirements and a strong AUP", which still had lots of stuff about dial-up users and such that didn't apply to me. And is "I find it extremely useful, from a diagnostic standpoint, to be able to review my SMTP delivery logs" a sufficient "operational requirement?" Like, when my wife's mail doesn't get delivered, I like to be able to tell her why?
The tone of the DUL pages, taken in total, is quite hostile -- or at a minimum paternalistic and condescending -- especially to individuals. Taken as a whole, it presents the attitude that individuals really don't need to have that kind of control over their Internet presence, that individuals should just trust in their ISPs and not worry thier little heads over it.
But, taking you at your word, I'll go ahead and make a request.
I'd be ecstatic if the DUL would go away. At a minimum, someone paying for a static IP address should be able to do direct SMTPs. And yes, I know the arguments, and no, I don't agree with them. Fix the protocol, don't dump on individuals.
Yes, but then that page at Red Hill is something of a relic. The VXpro rant was written about two years ago, while the rest of the page was written one year ago, in October of 1998. The link that they give to a rant at Carl Industries is dead -- Carl Industries has been shuttered since July of 1999. I recall, however, that even the page at Carl Industries had been softened somewhat before it disappeared; the author said that, to his surprise, he found PCchips to actually be responsive to the problems that their OEM customers faced, and that the quality had markedly improved since he first wrote those words.
I think that a lot of people have had bad experiences with PCchips boards some time ago, or with old boards sold recently and have (with good reason, clearly) stayed away from PCchips boards since. Just like with people who once had a bad experience with a certain brand of car, however, such people will bad-mouth the manufacturer for the rest of their lives without ever trying out the latest models to see if anything has changed.
Just about all boards -- even Tyan and Asus models; heck, even Sun Microsystems boards -- have non-zero early failure rates (AKA "Infant Mortality Rate" or "IMR"), and you can expect that a brand such as PCchips that is cutting things a hair's breadth from the edge will have higher rates than others. A small-time PC assembler that is leveraging low labor costs somewhere can afford to suffer that failure rate as long as the aggregate cost of all the boards that he has to buy to get the required number of good boards, together with the labor costs to do the testing and rework, is substantially less than what it would cost him to buy boards with significantly lower IMRs. If you compare a $40 PCchips board with a $100 Tyan board, it is likely that the assembler will only have to get around 50% good boards to come out ahead, and he probably gets much better than that -- I'd expect that better than 90% of PCchips boards work OK.
As an individual buying a single PCchips board, however, you're entering into something of a crap shoot. Still, as long as you (a) get at least a 90-day warranty, (b) understand the risk, and (c) view price as being much more important than performance, you will probably come out ahead with something like a PCchips board. It's all a matter of trade-offs
Although they aren't my favorites, I've used a couple of PCchips mobos, a VXpro model that my wife uses with a K5-200 (yes, K5; a few were actually shipped) to run WinNT 4.0, and a TXpro-II (SIS 5598 w/built-in VGA) that I've used to run everything from DR-DOS to Slackware 3.6. The XFree SIS server worked fine, although it was (duh) a little sluggish. As $50 motherboards with built-in SVGA and sound go, I'd say that they can't be beat.;-)
However, What I would hope to come out of this is that PCchips may actually start to pay attention to whether or not Linux runs on their motherboards; it would be kind of embarassing if they shipped the boards with an OS that didn't run on them (yeah, I know that they haven't seemed to have all that much shame in the past). Still, from this perspective, this deal probably offers about as much hope as we have any right to expect that Linux can be expected to run reliably on some of the cheapest & nastiest little PCs available on the market -- hell, maybe even the sound will work. This is not a bad thing; heretofore, the only thing they would have cared about would have been a basic level of Windows compatibility.
They're up to 3.0.3 on the release level now, and it's a pretty big improvement over 2.5. Some of the emMicro ports could use some cleaning up, and they're doing some of that now. I'm expecting a new PIC port soon, because the one that shipped with 3.0.3 didn't really fully support RS-485. Also, the C implementation of emMicro has cool things like gotos out of then clauses nested two deep.
However, they've changed their pricing and now all you can really buy is an unlimited OEM development and distribution license, and thus the price is about the same as that of a very nice automobile.
I agree that calling emMicro a "Web Server" is pretty lame, although it does provide "services" to emGateway.
As to whether there are any shipping products, it's hard to say. They have a lot of press releases disclosing design wins, but those tend to be of the "these people say they're gonna use our stuff for this cool new product they're working on" sort, while OEMs tend not to be so interested in blowing emWare's horn when the product ships.
My project is in-house only (and I can only really afford to do so because I bought before they changed the pricing structure), so it will never "ship" in any traditional meaning of that word.
I (or my office, more properly) have a development license for emWare, and I've built a couple of devices that work with it. Basically, they provide some Win32 DLLs that will front end some low-overhead back-end networks: RS-232, multidrop/full-duplex RS-485, modem... also ethernet is supported, and I think that there's an X10 port of the protocol. The microcontrollers run a fairly light-weight software called emMicro; it is coded as an event loop that, each time through, does basic housekeeping -- doing all the stuff that microcontrollers do -- but also checking for requests coming in on the network. The protocol provides for the Win32 host to discover the capabilities of the controller, and then to access those capabilities over the link; on that level it works sort of like an ORB or portmapper.
On the Win32 side, they provide a special-purpose proxy service that, on the one side, talks http to a TCP/IP network, and on the other it talks to the back end network (called emNet -- what else) to service requests coming in from the TCP/IP network. One could probably do the same sort of thing by devising a lightweight protocol to talk over a dedicated network of microcontrollers, and then writing service routines that were callable by CGI scripts off of Apache (that's what I was going to do before I found this software -- emWare probably took the better part of a year off my development schedule). Nevertheless, that would be a bunch of work, and emWare's already done it. Main downside is that it only runs on Win32 (95/58, NT & CE). The thing I like most about it is that you can focus on what the controller is supposed to be doing, and you don't have to screw around with matrix keypads and LCD displays and such. Once you get over the big hump in the learning curve, adding new functions is a breeze.
In addition to the http interface, they also have some Java Bean controls, and an API callable from C++.
I have it working on PICs (16C76) and some Hitachi H8 units (the H8S/2134 boards that emWare sells), and I'm working on an 8051 design now. I'm using multidrop RS-485. It's pretty straightforward once you see what it's doing, but it is quite expensive to get started with more than their eval license.
The one thing I notice on these threads is the criticisms of Raster for being immature. This is like criticizing a just-born kitten for having its eyes closed. How old is Raster, anyway (I thought he was in his early twenties, but I could of course be wrong), and from where was he supposed to get all that wonderful maturity everyone thinks he should have? If someone possessing as much raw talent and in receipt of as much recognition as Raster can't be a little hot-headed at that age, then I think that we've adopted a set of expectations for human behavior that is entirely disconnected with reality.
If Raster's departure screws up anything in Red Hat's plans, then it is arguably true that Red Hat has some rather short-sighted managers; placing that great a dependence on someone as young and mercurial as Raster would a classic lapse of common business sense. I would expect that they hired Raster as someone with a deep understanding of X and an uncanny ability with graphical textures, not as someone who would carry the battle flag to victory. I'm sure that others working there have learned a great deal from him, that many of his contributions are greatly appreciated, and that the work will continue without him.
Raster is strong-willed and talented enough (even if he can't write to save his life) that he will probably go on to do great things, whether it is by himself or leading some others. I am also confident that, to the extent that his behavior in this case reflects immaturity, some of that immaturity will be worn off as he experiences the consequences thereof. I wish him well.
I think that the introduction of restrictions on links to non-free content on/. would be a bad idea. I think that it is valuable for one to know that IEEE Spectrum published such an article, even if one can't read it online. If Rob and/or someone else in the crew thinks it's worth putting up, then fine. The/. item was very explicit that the Spectrum article wasn't available to non-IEEE members, so no one should have been surprised when they found that there wasn't any way around it.
It's not like there isn't any filter on/. content... I've submitted maybe a dozen stories and not a one has been picked up (except maybe one or two that was put up when someone else submitted them a couple of weeks later... I'm sure everyone who submits stories has had this happen).
BTW, the IEEE is doing much better on keeping up with current events in their publications; the print journals are still dismally behind the times, but they have taken to publishing much more stuff online, and the membership includes good access even to stuff you don't subscribe to (subscribing to one journal can get you access to the online stuff in some other journals).
According to the latest (5/10/99) Microprocessor Report, Compaq has a fairly aggresive roadmap for Alpha. The item, "Alpha Roadmap Gets Clearer" appears in the "Most Significant Bits" column on page 4 and is attributed to Linley Gwennap. It cites "Long-time Alpha Watcher" Terry Shannon among others. Here's a very quick summary of what the story says:
0.25u 21264 (EV67) @ 800MHz by year-end.
First 700MHz EV67 parts are expected to spec out at 30 SPECint95/base, 60 SPECfp95/base.
0.18u 21264 (EV68) @ 1GHz in 1Q00, spec at 50 int/85 fp.
Hmmm. Well, now I had written to Rob with a suggestion for a new Slashbox with a button that would just automatically pick the "Rob sucks" selection on whatever the poll happened to be that day. I don't know, but perhaps this is his way of saying "don't count on it"...:-)
But I agree... the new code is a huge improvement.
CheckFree, which I've used for years and, others' problems notwithstanding, I've been extremely happy with (they've even written letters on my behalf to idiot creditors) sells a web-based service that will work with any bank. There's an enrollment form here. The really cool thing about using CheckFree is being able to sit down once or twice a month and schedule bills to be paid exactly on the due date... no more sending checks early because you don't want to be bothered with it later.
Prob. not free like much of SuSE. Stick w Beowulf!
on
SuSE Cluster
·
· Score: 1
In my experience, the only *pre-selected* "non-free" package is xv, and they do in fact point out that it has a restrictive, shareware-type license. I also recall that, when you select the qt libraries, there is a pop-up cautioning you to make sure that you understand the license terms for qt. It always seemed to me that they were simply less dogmatic than RedHat. My biggest problem with SuSE was always that, while they include an astounding number of packages, they don't seem to be able to keep up with recent releases. Some of the stuff on SuSE 5.3 is over a year old and three versions behind. Since they don't have the same level of contrib activity as RedHat, one often has to get down and dirty with RPM SPEC files to bring your system up to date. That, and the fact that it was virtually impossible to install a 2.1 kernal without just about rebuilding the whole OS. I really hope that they've fixed most of these problems in 6.0, which I haven't tried yet.
Check out JK Micro's FlashTCP Embedded Web Server . Remember that these things are designed to simply feed monitoring info up onto the net; they're not designed to be computing powerhouses...
I've been trying to get drzyzgula.org moved off of granitecanyon.com (which was for the most part great when it was up but I find as I depend more on it I want to get commercial-level service) and on to my ISP's (eskimo.com) DNS servers for a week. My ISP has sent in the request twice and all we've gotten back is the autoreply. I sent it in again yesterday and have heard nothing. One week, three submissions, one automatic reply, no ACK requests: (*@? @#$ (@ $@( @#$ (*@Y!@$ @# $@. At least my old DNS servers are still working, so I can get email. They *used to* have a system where you could enter the tracking number of your request in a web form and look up the status of the request. I notice that that system is gone gone gone, and no wonder. BAH!
Learning curve. There's nothing like getting a desktop that does approximately nothing to lure you into any sense of comfort. I'm sure that it is straightforward to add menus, but I had trouble figuring out how get myself functional enough to read what documentation there was. Even twm in X classic gives you more help here. The documentation that RedHat is writing should help. Also, starting up the Gnome panel by default will help as well, since that comes with a bunch of pre-done menus. However, when I started up the 0.99 build, clicking on the start menu scattered stacked menu trees all over the place.
Icon management. This could well be a learning curve thing, but I found I had to place icons manually, which is a real drag. I still pine for twm's icon managers, which give nice, compact and unobtrusive button arrays organized by application.
Intrusiveness. I did a fresh build of Red Hat 5.2, including most stuff except for the IPX stuff and any form of emacs (ugh!). I then grabbed the 0.99 Gnome release. In order to install gnome, I had to back out:
AfterStep-APPS-1.5-0.3
control-panel-3.7-7
gecko-1.6-0
gimp-1.0.1-2
gnome-linuxconf-0.14-4rh
usermode-1.4.3-1
usernet-1.0.8-1
wmakerconf-1.1.1-3
AfterStep-1.5-0.7
printtool-3.29-3
timetool-2.3-7
(I saved the list so I could un-do everything). This seems a bit much, especially since there isn't any obvious way to restore the functionality.
The one other comment I'd like to make about E, and I could certainly RTFM (it could be there and I just don't know it) but it is a point that anyone considering developing a window manager might take into account so I'll say it here: Please give a good amount of macro processing on the window manager's configuration file. At my office, we support about 350 users in a production X environment. We need to be able to deliver.*wmrc configuration files that (a) make sure that everyone has menus that deliver all the basic software, (b) allow for extensive per-user and per-workgroups customization and (c) are rdistable. This is extraordinarilly difficult to provide simply by giving include-file hooks. We still make extensive use, for example, of tvtwm and ctwm, precisely because of the extensive support for m4 preprocessing of the configuration files.
But this is intended only as constructive criticism. E is an amazing piece of work, and the separation from the Gnome panel so that other, simpler wms can be used is a huge advantage of Gnome over other systems. No one should be flamed for developing a functional piece of free software. RedHat cannot shove anything down anyone's throat. Even when RH 6.0 comes out, assuming that it has a full-blown Gnome/E environment by default, users will still have a choice in desktop software, which those of us with memories going back to the '70s and '80s can only see as a wonderful, novel development.
Slashdot Mirror
A follow-up here for the record. I put in a request, and it was granted almost immediately; the DUL guys poked around a bit and when it checked out they went ahead and took my IP out of the DUL jail. I take back what I said. Well, most of it anyway, except for the part about the tone of their Web page, which I still think sounds pretty hostile. Nice to know, though, that they're more reasonable than they sound. --Bob
Well, that's encouraging. I did not do so because after reading through all of the information at http://maps.vix.com/dul/, I could find no indication that such a request would be honored. On the page of information for end users, http://maps.vix.com/dul/enduser.htm, the exclusive remedy is to use your ISP's relay. On the page on removing your network, http://maps.vix.com/dul/removing.htm, the acceptable reasons listed speak only to the needs of ISPs, not individuals. The closest thing I could find was the clause on "Removal due to operational requirements and a strong AUP", which still had lots of stuff about dial-up users and such that didn't apply to me. And is "I find it extremely useful, from a diagnostic standpoint, to be able to review my SMTP delivery logs" a sufficient "operational requirement?" Like, when my wife's mail doesn't get delivered, I like to be able to tell her why?
The tone of the DUL pages, taken in total, is quite hostile -- or at a minimum paternalistic and condescending -- especially to individuals. Taken as a whole, it presents the attitude that individuals really don't need to have that kind of control over their Internet presence, that individuals should just trust in their ISPs and not worry thier little heads over it.
But, taking you at your word, I'll go ahead and make a request.
I'd be ecstatic if the DUL would go away. At a minimum, someone paying for a static IP address should be able to do direct SMTPs. And yes, I know the arguments, and no, I don't agree with them. Fix the protocol, don't dump on individuals.
--Bob
Yes, but then that page at Red Hill is something of a relic. The VXpro rant was written about two years ago, while the rest of the page was written one year ago, in October of 1998. The link that they give to a rant at Carl Industries is dead -- Carl Industries has been shuttered since July of 1999. I recall, however, that even the page at Carl Industries had been softened somewhat before it disappeared; the author said that, to his surprise, he found PCchips to actually be responsive to the problems that their OEM customers faced, and that the quality had markedly improved since he first wrote those words.
I think that a lot of people have had bad experiences with PCchips boards some time ago, or with old boards sold recently and have (with good reason, clearly) stayed away from PCchips boards since. Just like with people who once had a bad experience with a certain brand of car, however, such people will bad-mouth the manufacturer for the rest of their lives without ever trying out the latest models to see if anything has changed.
Just about all boards -- even Tyan and Asus models; heck, even Sun Microsystems boards -- have non-zero early failure rates (AKA "Infant Mortality Rate" or "IMR"), and you can expect that a brand such as PCchips that is cutting things a hair's breadth from the edge will have higher rates than others. A small-time PC assembler that is leveraging low labor costs somewhere can afford to suffer that failure rate as long as the aggregate cost of all the boards that he has to buy to get the required number of good boards, together with the labor costs to do the testing and rework, is substantially less than what it would cost him to buy boards with significantly lower IMRs. If you compare a $40 PCchips board with a $100 Tyan board, it is likely that the assembler will only have to get around 50% good boards to come out ahead, and he probably gets much better than that -- I'd expect that better than 90% of PCchips boards work OK.
As an individual buying a single PCchips board, however, you're entering into something of a crap shoot. Still, as long as you (a) get at least a 90-day warranty, (b) understand the risk, and (c) view price as being much more important than performance, you will probably come out ahead with something like a PCchips board. It's all a matter of trade-offs
Although they aren't my favorites, I've used a couple of PCchips mobos, a VXpro model that my wife uses with a K5-200 (yes, K5; a few were actually shipped) to run WinNT 4.0, and a TXpro-II (SIS 5598 w/built-in VGA) that I've used to run everything from DR-DOS to Slackware 3.6. The XFree SIS server worked fine, although it was (duh) a little sluggish. As $50 motherboards with built-in SVGA and sound go, I'd say that they can't be beat. ;-)
However, What I would hope to come out of this is that PCchips may actually start to pay attention to whether or not Linux runs on their motherboards; it would be kind of embarassing if they shipped the boards with an OS that didn't run on them (yeah, I know that they haven't seemed to have all that much shame in the past). Still, from this perspective, this deal probably offers about as much hope as we have any right to expect that Linux can be expected to run reliably on some of the cheapest & nastiest little PCs available on the market -- hell, maybe even the sound will work. This is not a bad thing; heretofore, the only thing they would have cared about would have been a basic level of Windows compatibility.
--BobHowever, they've changed their pricing and now all you can really buy is an unlimited OEM development and distribution license, and thus the price is about the same as that of a very nice automobile.
I agree that calling emMicro a "Web Server" is pretty lame, although it does provide "services" to emGateway.
As to whether there are any shipping products, it's hard to say. They have a lot of press releases disclosing design wins, but those tend to be of the "these people say they're gonna use our stuff for this cool new product they're working on" sort, while OEMs tend not to be so interested in blowing emWare's horn when the product ships.
My project is in-house only (and I can only really afford to do so because I bought before they changed the pricing structure), so it will never "ship" in any traditional meaning of that word.
On the Win32 side, they provide a special-purpose proxy service that, on the one side, talks http to a TCP/IP network, and on the other it talks to the back end network (called emNet -- what else) to service requests coming in from the TCP/IP network. One could probably do the same sort of thing by devising a lightweight protocol to talk over a dedicated network of microcontrollers, and then writing service routines that were callable by CGI scripts off of Apache (that's what I was going to do before I found this software -- emWare probably took the better part of a year off my development schedule). Nevertheless, that would be a bunch of work, and emWare's already done it. Main downside is that it only runs on Win32 (95/58, NT & CE). The thing I like most about it is that you can focus on what the controller is supposed to be doing, and you don't have to screw around with matrix keypads and LCD displays and such. Once you get over the big hump in the learning curve, adding new functions is a breeze.
In addition to the http interface, they also have some Java Bean controls, and an API callable from C++.
I have it working on PICs (16C76) and some Hitachi H8 units (the H8S/2134 boards that emWare sells), and I'm working on an 8051 design now. I'm using multidrop RS-485. It's pretty straightforward once you see what it's doing, but it is quite expensive to get started with more than their eval license.
The one thing I notice on these threads is the criticisms of Raster for being immature. This is like criticizing a just-born kitten for having its eyes closed. How old is Raster, anyway (I thought he was in his early twenties, but I could of course be wrong), and from where was he supposed to get all that wonderful maturity everyone thinks he should have? If someone possessing as much raw talent and in receipt of as much recognition as Raster can't be a little hot-headed at that age, then I think that we've adopted a set of expectations for human behavior that is entirely disconnected with reality.
If Raster's departure screws up anything in Red Hat's plans, then it is arguably true that Red Hat has some rather short-sighted managers; placing that great a dependence on someone as young and mercurial as Raster would a classic lapse of common business sense. I would expect that they hired Raster as someone with a deep understanding of X and an uncanny ability with graphical textures, not as someone who would carry the battle flag to victory. I'm sure that others working there have learned a great deal from him, that many of his contributions are greatly appreciated, and that the work will continue without him.
Raster is strong-willed and talented enough (even if he can't write to save his life) that he will probably go on to do great things, whether it is by himself or leading some others. I am also confident that, to the extent that his behavior in this case reflects immaturity, some of that immaturity will be worn off as he experiences the consequences thereof. I wish him well.
I think that the introduction of restrictions on links to non-free content on /. would be a bad idea. I think that it is valuable for one to know that IEEE Spectrum published such an article, even if one can't read it online. If Rob and/or someone else in the crew thinks it's worth putting up, then fine. The /. item was very explicit that the Spectrum article wasn't available to non-IEEE members, so no one should have been surprised when they found that there wasn't any way around it.
It's not like there isn't any filter on /. content... I've submitted maybe a dozen stories and not a one has been picked up (except maybe one or two that was put up when someone else submitted them a couple of weeks later... I'm sure everyone who submits stories has had this happen).
BTW, the IEEE is doing much better on keeping up with current events in their publications; the print journals are still dismally behind the times, but they have taken to publishing much more stuff online, and the membership includes good access even to stuff you don't subscribe to (subscribing to one journal can get you access to the online stuff in some other journals).
- 0.25u 21264 (EV67) @ 800MHz by year-end.
- First 700MHz EV67 parts are expected to spec out at 30 SPECint95/base, 60 SPECfp95/base.
- 0.18u 21264 (EV68) @ 1GHz in 1Q00, spec at 50 int/85 fp.
- 0.18u 21364 (EV7) @ 1GHz by 4Q00-1Q01, expected spec 60 int/100 fp.
- EV8 expected 2002.
- EV9 and EV10 are in development.
That doesn't sound like dropping Alpha to me.Hmmm. Well, now I had written to Rob with a suggestion for a new Slashbox with a button that would just automatically pick the "Rob sucks" selection on whatever the poll happened to be that day. I don't know, but perhaps this is his way of saying "don't count on it"... :-)
But I agree... the new code is a huge improvement.
CheckFree, which I've used for years and, others' problems notwithstanding, I've been extremely happy with (they've even written letters on my behalf to idiot creditors) sells a web-based service that will work with any bank. There's an enrollment form here. The really cool thing about using CheckFree is being able to sit down once or twice a month and schedule bills to be paid exactly on the due date... no more sending checks early because you don't want to be bothered with it later.
In my experience, the only *pre-selected* "non-free" package is xv, and they do in fact point out that it has a restrictive, shareware-type license. I also recall that, when you select the qt libraries, there is a pop-up cautioning you to make sure that you understand the license terms for qt. It always seemed to me that they were simply less dogmatic than RedHat. My biggest problem with SuSE was always that, while they include an astounding number of packages, they don't seem to be able to keep up with recent releases. Some of the stuff on SuSE 5.3 is over a year old and three versions behind. Since they don't have the same level of contrib activity as RedHat, one often has to get down and dirty with RPM SPEC files to bring your system up to date. That, and the fact that it was virtually impossible to install a 2.1 kernal without just about rebuilding the whole OS. I really hope that they've fixed most of these problems in 6.0, which I haven't tried yet.
Check out JK Micro's FlashTCP Embedded Web Server . Remember that these things are designed to simply feed monitoring info up onto the net; they're not designed to be computing powerhouses...
I've been trying to get drzyzgula.org moved off of granitecanyon.com (which was for the most part great when it was up but I find as I depend more on it I want to get commercial-level service) and on to my ISP's (eskimo.com) DNS servers for a week. My ISP has sent in the request twice and all we've gotten back is the autoreply. I sent it in again yesterday and have heard nothing. One week, three submissions, one automatic reply, no ACK requests: (*@? @#$ (@ $@( @#$ (*@Y!@$ @# $@. At least my old DNS servers are still working, so I can get email. They *used to* have a system where you could enter the tracking number of your request in a web form and look up the status of the request. I notice that that system is gone gone gone, and no wonder. BAH!
- Learning curve. There's nothing like getting a desktop that does approximately nothing to lure you into any sense of comfort. I'm sure that it is straightforward to add menus, but I had trouble figuring out how get myself functional enough to read what documentation there was. Even twm in X classic gives you more help here. The documentation that RedHat is writing should help. Also, starting up the Gnome panel by default will help as well, since that comes with a bunch of pre-done menus. However, when I started up the 0.99 build, clicking on the start menu scattered stacked menu trees all over the place.
- Icon management. This could well be a learning curve thing, but I found I had to place icons manually, which is a real drag. I still pine for twm's icon managers, which give nice, compact and unobtrusive button arrays organized by application.
- Intrusiveness. I did a fresh build of Red Hat 5.2, including most stuff except for the IPX stuff and any form of emacs (ugh!). I then grabbed the 0.99 Gnome release. In order to install gnome, I had to back out:
The one other comment I'd like to make about E, and I could certainly RTFM (it could be there and I just don't know it) but it is a point that anyone considering developing a window manager might take into account so I'll say it here: Please give a good amount of macro processing on the window manager's configuration file. At my office, we support about 350 users in a production X environment. We need to be able to deliver- AfterStep-APPS-1.5-0.3
- control-panel-3.7-7
- gecko-1.6-0
- gimp-1.0.1-2
- gnome-linuxconf-0.14-4rh
- usermode-1.4.3-1
- usernet-1.0.8-1
- wmakerconf-1.1.1-3
- AfterStep-1.5-0.7
- printtool-3.29-3
- timetool-2.3-7
(I saved the list so I could un-do everything). This seems a bit much, especially since there isn't any obvious way to restore the functionality.But this is intended only as constructive criticism. E is an amazing piece of work, and the separation from the Gnome panel so that other, simpler wms can be used is a huge advantage of Gnome over other systems. No one should be flamed for developing a functional piece of free software. RedHat cannot shove anything down anyone's throat. Even when RH 6.0 comes out, assuming that it has a full-blown Gnome/E environment by default, users will still have a choice in desktop software, which those of us with memories going back to the '70s and '80s can only see as a wonderful, novel development.