First, be aware that DSL or cable internet service providers (ISP), economically, are interested in selling (or offering) USB and Wi-Fi technologies, but for your interest, you must ask for Ethernet technology only (even if that means that you have to wait longer for delivery, or pay more). This way you will get hardware compatibility with many systems (video games consoles, home appliances, new and rare operating systems, etc...), a much better reliability over long time, and incredible cheap cost for spare parts needed when a repair is needed. Below you will find details about Homeplug standard in order to avoid wires, even with Ethernet technology (to not want wires does not mean you are forced to select Wi-Fi, furthermore statistics show that companies use much more Homeplug than Wi-Fi). About speed choice (512K, 1024K, etc...), you must know that maximal server speed (bandwidth) is divided by the number of clients connected to it and the actual speed you will be able to observe will almost always be under 512K (512Kbits/s, 1 byte=8 bits, thus around 60Kb/s). Thus, select always the cheapest offer, often 512K, and if it works fast enough for you, keep it. You should test higher speeds only if you actually observe that your speed can't go beyond 60Kb/s when you download something (special case of a server that only handles you).
If Miniconf detects file and dial number, username, dns are valid, then it should work.
Untested PCMCIA network adapters :
You may request assistance through full support.
A Cardbus router DOS program is required and MUST exist for the adapter.
Minilgos will probably require a small change in order to use correctly the router.
Untested PCI or ISA network adapters :
You can test yourself the appropriate packet driver (only if it's a .COM file).
It will probably work if it accepts interrupt number as parameter without flag (ex:" 0x60").
Copy it in Miniconf subdirectory "unknown". It will be detected (but will disable pcmcia).
Then, launch Miniconf to configure your ISO image before burning your CD-ROM.
If it fails, do some test under DOS in order to retrieve error messages & use light support.
Unsupported USB network adapters :
At the moment, there is no plan to have Minilgos support USB adapters.
Existing drivers are too much OS-dependent.
If you want to use this unsupported adapter, you have to use it through a "gateway" (see above).
Unsupported Wi-Fi network adapters :
At the moment, there is no plan to have Minilgos support Wi-Fi adapters.
Existing drivers are too much OS-dependent. Not recommended for children environment.
For family wireless need we suggest powerline-ethernet adapters (www.homeplug.com.)
If you want to use this unsupported adapter, you have to use it through a "gateway" (see above).
About packet drivers :
Minilgos needs a "packet driver" in order to work with a network adapter. It fulfills an OS-independent driver specification created by "FTP Software, Inc." that needs almost only BIOS calls (and a few easy to emulate DOS calls) to work correctly. They are often located in directory "PKTDRV" on floppy disk or cdrom found with the network adapter inside the package.
PCI network adapters are the best. They have two ID numbers allowing a fast and reliable detection (Vendor_ID & Device_ID). These numbers appear on screen at BIOS startup time. Just check for a line ending with "Network Controller". Minilgos will detect these numbers and if they belong to a known successfully tested PCI network adapter, the appropriate packet driver will be immediately loaded in memory and launched (assuming Miniconf, the Minilgos cdrom image configuration utility could copy packet driver file inside the cdrom .ISO image). If you try Minilgos on a desktop computer that has a cheap and common network adapter, it should work instantly without any special change in Miniconf packet drivers directory (3rdparty).
If none of them could be detected, then Minilgos will attempt to launch a successfully tested packet driver for the PCMCIA network adapter SMC 8036TX, "smcpkt.com" with its CardBus router "smcopen.exe" . If you need another PCMCIA adapter, you will have to subscribe to full support and ask for it (a cardbus router will be required, and MUST exist). To support at least one PCMCIA adapter is important since laptop computers use PCMCIA instead of PCI for additional peripherals (but if you are lucky your laptop may have a PCI integrated network interface card on mother board and packet driver available). Minilgos does not support USB adapters (both USB ports and devices drivers are too much OS-dependent). So for now, Minilgos may work on your laptop but only with PCMCIA adapter SMC 8036TX or a PCI integrated adapter with available packet driver.
You can try yourself untested PCI or ISA or whatever network adapters. You just copy its .COM packet driver in Miniconf "unknown" directory. It will be detected and will replace the PCMCIA packet driver. Note that the packet driver you want to try MUST have a size below 60Kb. So far I never failed to find a packet driver for a PCI network adapter, so you should manage to find it (be sure to check web site of the appropriate company, but avoid sites of local subsidiaries).
If you try Wi-Fi adapters (if you ever find a packet driver), be aware that FTP and POP3 protocols usually requires uncrypted login password, often valid only with your specific ISP connection, but a skilled hacker may invade your "Air" local network from behind walls and read/write everything. You may also have radio-wave conflicts with neighbourhood (right when you try it, or maybe a few months later...). So, try Wi-Fi at your own risk, you have been warned. A cute alternative for wireless connexion would be to use your house power lines. Take a look at site www.homeplug.com and on french site www.courantmultimedia.fr for european power plugs. Price is around 60€ per powerline-ethernet interface, so you can divide your local network into two parts for 120€ (for example phone line and DSL/cable modem on one side downstairs, and your hub and family local network on the other side upstairs) without being vulnerable to radio-waves troubles and without the need to change your ethernet devices, assuming the limited 150 meters range is acceptable for you. Only issue I know, not a serious one, is sometimes the need to unplug/replug adapter when bandwidth falls down completely, in order to force it to readapt itself to the electrical waveform because this waveform was not stable over time in your house and changed slightly (may happen around once a week or never, depending on your electrical installation quality). Price may appear expensive to you but Wi-Fi may require that you replace your DSL/cable modem and ALL your ethernet OS-independent adapters with OS-dependent Wi-Fi adapters so that may be also expensive, quite nasty since you get instantly married to only one OS and risky if it fails the day your neighbourhood decides to use Wi-Fi as well...
Minilgos author has also a personal reason for avoiding Wi-Fi technology, especially when children are involved, but this reason is not currently proved at the moment (but may be proved in future, after many years maybe). An electro-magnetic signal energy goes down according to a 1/(R*R) function where R is distance between you and transmitter. That means that we should not be afraid about regional TV or Radio transmitters which are probably a few miles away from you (TV and radio tuners, because of these great distances, need very powerful amplifiers to recover signal). We know, because many official health administration discourage it (for example, the Stewart report - see paragraph 1.53 in report summary), that we should not let children use mobile phones (it seems that below age of 16 years old, electro-magnetic signal penetrates far more deeply inside body, so, much more cells get exposed to significant energy, and also, periodic minor damages perturbate a perfect growth and development). Brain damage was observed on mouses, but also humans after long time phone calls with a short distance (a few inches, GSM signal with frequency 0.9Ghz or 1.8Ghz) between brain and mobile phone transmitter (cancer is not proved even with these observed damages, since most damage creates only headaches and repairs itself over time and cancer is a very rare event where damage creates an abnormal, but valid, genetic "program" inside a cell). Problem with this kind of danger is that it's not a predictable poison affecting immediately a person, it's more likely related to probabilities. Further more it's an invisible danger able to go through walls and that can be stopped only by conductors (metal or water for example). We know that UV sun rays damage DNA branches, and we know that we have enzyms that rebuild them (molecules in DNA branches are doubled, if half of it is broken, enzyms recognize other half and use it to find out what molecule needs to be re-installed in the DNA branch), and in case of extreme sun damage, skin cells are unable to repair it fast enough and you get accelerated skin withering and significant skin cancer probability, even several years later if affected gene is sleeping (cancer is actually, in this case, a probability matter, i.e how many chance you get to have a rare and specific bad interaction between electro-magnetic signal, or molecule, and a cell genetic "program", coded in DNA).
So, Minilgos author believes that there are two simultaneous conditions to avoid, in order to keep cancer probability as low as possible. First condition to avoid is to use systems with local electro-magnetic transmitters very close to you or your children. Second condition to avoid is to use systems working for very long time. You should avoid at all cost, to fulfill simultaneously both conditions, but separately you can probably afford them. The distance to transmitter will define how low cancer probability is for a short duration (depending on signal energy, exact distance, frequencies -but safe frequencies probably do not exist-, 2.4Ghz for Wi-Fi 802.11b/802.11g or 5Ghz for Wi-Fi 802.11a, for example micro wave ovens use frequencies up to 2.45Ghz and have great thermal impact on water H2O molecules-), and the actual exposure duration will multiply this low probability (in fact probabilty won't raise strictly as a linear function, but it's quite linear as soon as exposure starts, as shown in graph below). The clue is to not get a multiplication result that gives a significant cancer probability...
For example, for radioactive signals, energy is so high, because of so high frequency, that calculation of the total amount of exposure time a human can afford during lifetime without risking too much to catch a cancer can be reliably estimated -a few minutes- because of past experiences and special badges were designed to give a visible colored alert when received energy multiplied by exposure time exceeds a specific amount; another example is X-ray transmitters in hospitals : hospital accepts that you receive X-rays for a very short duration, only a few times in your entire life, but will never accept that people in charge of this machine get exposed even once because repeating exposure time multiplies cancer probability, so they protect themselves behind a lead glass and a wall; finally, a last terrific example : officialy, armies using low power uranium ammunitions, bombs or missiles -more effective because uranium is heavier, penetrates steel easily and is low cost since nuclear industry is quite happy to throw away this part of nuclear wastes- declare they can prove any time that soldiers will never catch a cancer when they use them -by the way, uranium, even without radio-activity, from a chemical point of view, is as toxic as arsenic when it reaches environment, water, etc..., but it's another story-. They can prove it because exposure time is short. On the other hand, statistics show that children raised in territories where these ammunitions were used massively often catch cancer after an average of ten years. More precisely, a TV reporter explained that 1 out of 10 children in Iraqi dies because of blood cancer, today, near first gulf war battle zones. If you want to hear what she said by yourself here are the references : Broadcasted on french TV channel "France 3"; Reporter name is Caroline SINZ, with Christian de CARNE, Salah AGRABI, Michelle GUILLOISEAU-JOUBAIR; Title was "Hopital pediatrique de Bagdad"; Date was November the 6th in 2004. So, that means very long exposure time to low energy is the same as short exposure time to high energy. Of course radio-active rays have much higher energy than standard radio-wave rays but it's quite interesting to study them and proves the link between exposure time and final cancer probability.
So, unfortunately, Wi-Fi would force children to stand very close to transmitter (Wi-Fi network card), activated almost permanently, because internet connections are, by nature, almost permanents nowadays (we all use computers every day for hours and it seems that TV will also use Wi-Fi, so we can consider Wi-Fi signal will be permanent in your house). Experts may tell you that Wi-Fi signal energy is much lower than mobile phones signal energy for example, and they are right, that lowers enormously cancer probability (for same exposure time), but they forget that repeating exposure time raises enormously the same probability as well (and can, thus, compensate signal energy difference). So, why science can't (nowadays) prove such danger, is probably because its probability is the result of a multiplication between two numbers, one very low, and another very high (related to very long duration)... At the moment, humanity (any government) does not have the money nor the time to perform such calculations for each electronic device, for each usage... Keep in mind, for example, that a scientist may offer to prove that lottery players never win, and observe for one year (obviously not enough), many players in a laboratory and effectively declares that no player did win. So you have to calculate risk yourself and decide what risk you accept. For minilgos, dedicated to children, Wi-Fi (unproved) related risks are not accepted, and ethernet link, eventually through homeplug technology, is recommended (minilgos may support USB devices in future, but the lack of independant power supply and the specific high current data format used along USB cable prevent this technology to be electrically secure and OS independant). At least, without Wi-Fi, parents willing to reduce electro-magnetic signals energy in their house will be able to do so, since enhanced cables exist for this purpose (ethernet, power, etc...) and metal or water can be used to block signals coming from outside. Sorry for this long subjective list of arguments, but for children, it was important to spend some time thinking deeply about this specific problem of repeating exposure time to electro-magnetic energy almost forever... Below you will find a short mathematical explanation :
This graph shows how a very low probability of catching a cancer, p=0.01 (means 1 chance out of 100 during exposure time T while exposed to a specific signal or molecule) raises when the exposure time is repeated (1*T up to 300*T). Probability of not catching it globally is multiplication of probabilities of not catching it in each small time period (1-p). As you can see the beginning of the function is quite linear, and very quickly probability becomes unacceptable, so exposure time MUST be as short as possible, even if the probability at beginning (calculated for a short exposure time) is very low.
Minilgos has also the ability to use a special packet driver named "PPP for DOS 0.6" by Antonio Lopez Molero (assuming Miniconf could copy the packet driver file into your .iso image -dosppp06.zip /chapsupp.zip /epppd.exe-). It allows Minilgos to connect to internet through a standard Hayes compatible modem connected with Com1 or Com2 (computer-modem speed set at 57600 bit per second, average internet-modem speed around 44000 bauds depending on phone line quality). This option is disabled if you configure an empty dial number or connection username with miniconf. DSL/cable users may be interested in configuring this option for the rare days where DSL/cable connection is out of order (happens...). But be aware this way to run Minilgos is very slow and sometimes you will need many tries to successfully connect... So it's supplied only for emergency backup or if you are traveling far away from any of your ISP access points. A good alternative would be to locate a cybercafe offering Minilgos compatible computers AND your ISP physical link -this way you enter, power on computer, insert your cdrom, restart computer to boot cdrom, remove cdrom once connected, access your mailbox and all your private documents, power off computer, pay your connection time, leave-.
Minilgos will first try to detect a Hayes modem on Com1/Com2, then any ethernet network adapter packet driver for PPPoE connection, then for a direct local, DHCP or BOOTP connection. So if you want to skip this first step be sure to have a blank dial number or your Hayes compatible modem turned off or disconnected from Com1/Com2.
Even if PPP over Ethernet packet drivers exist (for DSL or ADSL users, it's the same, ADSL is just DSL with upload and download bandwidth slightly different), be aware that you DON'T need them. Minilgos core has its own PPPoE (with CHAP authenticating method and MD5 encryption), DHCP & BOOTP client layer.
Sharing same account, all at same time through a "gateway" computer, instead of one at a time with a direct access, will of course slow down some operations, but not too much since Minilgos does small login/do-it/logout FTP actions allowing easily interlaced actions (FTP logins are exclusive and need a few minutes to reset if you disconnect suddenly and not correctly), but the main concern is POP & FTP storage capacities. Minilgos shows you a nice gauge to check free space (if your FTP server returns it in code 226 messages) but no more info. It would be too much time consuming to identify each used space owner... so if you need to save a lots of data in your FTP subdirectory you will have to use diplomacy and negociate with other users since you share with them the same FTP space... This is of course not a problem in a company intranet environment where you can ask much more FTP storage space). Usually ISP offers at least 10Mb FTP storage space and 10Mb POP storage space (FTP storage space is often inactive by default, you pay for it anyway but you have to manually activate it -it's often called "create your personal web page"- on your ISP home page before you can access it). Don't worry about the lack of FTP storage space, you will be able to create shortcuts to many additional free 100Mb POP3 storage space offered by companies like www.hotpop.com. Unlimited 512K DSL or ADSL or cable subscription is of course suggested when you share the same account (between 20 and 30 € per month, interesting price since it covers entirely all family needs, internet phone included, without any time limit or distance limit).
